Nitrogen Cycling in the Past and in the Present Mediterranean Sea and
Arabian Sea – Implications from Stable Isotope Studies
Dissertation
Zur Erlangung des Doktorgrades der Naturwissenschaften im Department
Geowissenschaften der Universität Hamburg
vorgelegt von
Jürgen Henning Möbius
aus
Hamburg
Hamburg
2010
Als Dissertation angenommen vom Department Geowissenschaften der
Universität Hamburg
auf Grund der Gutachten von
Prof. Dr. Kay-Christian Emeis
und
Dr. Birgit Gaye
Hamburg, den 7. Mai 2010
Prof. Dr. Jürgen Oßenbrügge
Leiter des Department Geowissenschaften
Zusammenfassung
Das Atomverhältnis der stabilen Stickstoffisotope (R=15N/14N, ausgedrückt als δ15N) in festen
und gelösten Stickstoffverbindungen wird oft genutzt, um N-Quellen und -Umsatzprozesse in
aquatischen Systemen zu verfolgen. In dieser Arbeit verwende ich isotopische Methoden mit dem
Schwerpunkt auf δ15N, um einzelne Bereiche des rezenten N-Kreislaufs im östlichen Mittelmeer
genauer zu untersuchen (Kapitel 3). Des Weiteren überprüfe ich anhand von kombinierten δ15NDaten und der Aminosäurezusammensetzung (ein Indikator für den Abbaugrad des organischen
Materials) aus Sinkstoffen, Oberflächensedimenten und Sedimentkernen die Anwendbarkeit des
δ15N als paläozeanografisches Werkzeug im östlichen Mittelmeer und im Arabischen Meer
(Kapitel 4 und 5).
Das
äußerst
nährstoffarme
östliche
Mittelmeer
weist
in
fast
allen
reaktiven
Stickstoffkompartimenten (also allen anorganischen und organischen Formen von N außer N2)
ungewöhnlich niedrige δ15N Werte auf. Als Grund dafür wurde diazotrophe N-Fixierung
vermutet, die bisher allerdings kaum beobachtet wurde und die zudem wegen des
Nitratüberschusses im Tiefen- und Zwischenwasser (Nitrat zu Phosphat- Verhältnisse bis zu 28)
biogeochemisch wenig sinnvoll erscheint. Um diese Umstände weiter zu klären, wurden
isotopische Daten der wichtigsten reaktiven N-Spezies (NO3-, gelöster reduzierter N,
suspendierter partikulärer N und sinkender partikulärer N) während einer Probenahmekampagne
im Winter/beginnenden Frühjahr 2007 im östlichen Mittelmeer bestimmt und nun in Kapitel 3
diskutiert und modelliert. Dabei zeigen die Mess- und Modellergebnisse, dass die wichtigste
NO3--Quelle für den Nitratpool während der Probenahme höchstwahrscheinlich remineralisiertes
NO3-
ist,
welches
durch
die
Nitrifizierung
von
absinkendem,
im
durchmischten
Oberflächenwasser produzierten partikulärem N entstanden ist. Außerdem wird ein signifikanter
Beitrag von atmosphärisch eingetragenen, industriell in Europa emittierten Stickoxiden auf das
δ15N-Signal im Oberflächenwasser deutlich.
Die paläozeanografische Anwendung von δ15N als Proxy wird generell dadurch erschwert,
dass das ursprüngliche Signal aus den biologisch produktiven oberen Wasserschichten während
des Absinkens der Partikel und nach der Sedimentation eine frühdiagenetische Überprägung
erfährt. Diesen frühdiagenetischen Effekt versuche ich in den Kapiteln 4 und 5 zu erfassen, um
das ursprüngliche isotopische Signal des in der euphotischen Zone assimilierten reaktiven N im
östlichen Mittelmeer (Kapitel 4) und im Arabischen Meer (Kapitel 5) während des Holozäns und
des späten Pleistozäns anhand von Sedimentkernen zu rekonstruieren. Dies ist durch eine
Kombination von δ15N-Werten mit der Aminosäurenzusammensetzung möglich, wobei letztere
genutzt wird, um den Abbaugrad des organischen Materials zu bestimmen. Die derart
i
normalisierten Sedimentkerndaten aus dem östlichen Mittelmeer zeigen, dass die großskaligen
(Jahrzehntausende umfassende) Veränderungen im δ15N-Signal und die horizontalen Gradienten
im δ15N von Sapropelen (isochrone, organisch-reiche Lagen) im wesentlichen durch
unterschiedliche Erhaltung aufgrund wechselnder Sauerstoffsättigung im Tiefen- bzw.
Bodenwasser bestimmt sind. Die erhöhten Akkumulationsraten von organischem Kohlenstoff, die
zur Bildung des jüngsten Sapropels (S1) führten, sind möglicherweise ebenfalls größtenteils auf
verhinderten Abbau unter sauerstoffdefizitären Bedingungen an der Sedimentoberfläche
zurückzuführen und damit weniger ein Ausdruck von erhöhter Primärproduktion.
Die großskaligen Schwankungen im δ15N-Signal der Kerne aus dem Arabischen Meer
(Kapitel 5) werden primär von Veränderungen im δ15N des assimilierten Nitrats hervorgerufen.
Hier wird die Variabilität
Monsunstärke
Zwischenwasser
und
damit
gesteuert;
hauptsächlich von externen klimatischen Veränderungen in der
einhergehenden
diese
unterschiedlichen
bestimmen
wiederum
die
Zirkulationsmustern
im
Sauerstoffgehalte
des
Zwischenwassers im Arabischen Meer und damit letztendlich die Isotopie des Nitrats. Die
Verteilungsmuster von δ15N in Oberflächensedimenten und in isochronen Zeitscheiben aus
Sedimentkernen sind hingegen als unterschiedliche Erhaltung des originalen Signals aus der
Primärproduktion und Nitratassimilation zu deuten.
Die Evaluierung von zwei Abbauindizes auf Aminosäurebasis (Reactivity Index, RI und
Degradation Index, DI) in rezenten Sedimenten und Sedimentkernen aus dem Arabischen Meer
legt nahe, dass der RI eher geeignet ist, den Erhaltungsgrad von organischem Material in
Sedimentkernen anzuzeigen als der DI. Durch die eindeutige Beziehung zwischen δ15N und dem
RI ist eine Rekonstruktion des originalen δ15N vor der diagenetischen Überprägung möglich.
ii
Abstract
The 15N/14N isotope ratio (δ15N) of particulate and dissolved N-species is widely used to trace
N sources and turnover processes in aquatic environments. In this thesis, I apply isotopic methods
with a focus on
15
N/14N ratios to elucidate the recent N-cycle in the Eastern Mediterranean Sea
(EMS) (Chapter 3). I further use data on the isotopic composition of N in conjunction with amino
acid composition (reflecting the state of degradation of organic matter) of sinking particulate
nitrogen (sediment traps), surface sediments and sediment cores to gauge the use of δ15N as a
paleoceanographic tool in recent sediments and sediment cores of the EMS and the Arabian Sea
(Chapters 4 and 5).
Unusually low δ15N values in many compartments of reactive N in the extremely nutrient poor
EMS have been attributed to diazotrophic N2 fixation - although observational evidence is scarce
and the surplus of nitrate over phosphate (N to P ratios up to 28) in deep and intermediate waters
argue against the biogeochemical need of N2 fixation. Isotope data raised on major N species
involved in the cycling of reactive N (NO3-, total dissolved reduced N, suspended particulate N
and sinking particulate N) in the EMS during winter/early spring of 2007 are discussed and
modeled in Chapter 3. Analytical and model results suggest that recycled NO3- from nitrification
of sinking particulate N produced in the surface mixed layer is the most likely source of NO3during the winter sampling. Additionally, a strong imprint of atmospherically deposited nitrous
oxides of industrial origin in Europe on the δ15N of the mixed layer NO3- is indicated.
The paleoceanographic application of δ15N as a proxy is generally hampered by an early
diagenetic overprint of the signal formed in the ocean surface mixed layer that occurs during
particle sinking and after sedimentation. In Chapter 4 and 5, I quantify the effect of early
diagenetic alterations on sedimentary N-isotope composition in order to reconstruct the pristine
signal of N assimilated in the euphotic zone of the EMS (Chapter 4) and the Arabian Sea (Chapter
5) during the Holocene and late Pleistocene. This is possible by using paired data of δ15N and
amino acid composition, which can be applied to estimate the degree of organic matter
degradation. The normalized records for the EMS reveal that millennial scale oscillations in δ15N,
and its spatial gradients in organic rich layers (sapropels) are driven mainly by differences in
preservation under oxygen-poor and oxygen-replete bottom water conditions. The EMS study
gives further evidence that enhanced organic carbon accumulation in the youngest sapropel S1
may be due to better preservation under suboxic to anoxic conditions at the sediment-water
interface, and to a lesser extent to elevated productivity.
In the Arabian Sea δ15N records, millennial scale oscillations are mainly controlled by
changes of the δ15N of assimilated source nitrate; these changes are to a large extent caused by
iii
external climatic variability in the monsoon dynamics and associated circulation of the
intermediate water masses in the Arabian Sea, which in turn determine oxygen levels in midwater and nitrate isotopic composition. Most of the δ15N amplitude in spatial patterns of surface
sediments and isochronous time slices in sediment cores can be attributed to differences in
preservation of the original δ15N signal exported from the sea surface. An evaluation of two
preservation indexes based on amino acid composition (Reactivity Index, RI, and Degradation
Index, DI) in both recent sediments and core samples from the Arabian Sea suggest that the RI is
more suitable than the DI in estimating the state of organic matter degradation in core sediments
that have undergone progressive degradation. The relationship between δ15N and the RI allowed a
reconstruction of the original δ15N prior to diagenetic overprinting.
iv
Contents:
Zusammenfassung ________________________________________________________ i
Abstract _________________________________________________________________iii
Contents _________________________________________________________________ v
1. Introduction __________________________________________________________ 1
2. Working areas, projects and thesis outline _________________________________ 7
3. N-isotope ratios of nitrate, dissolved organic nitrogen and particulate nitrogen
in the Eastern Mediterranean Sea ________________________________________11
4. Diagenetic control of nitrogen isotope ratios in Holocene sapropels and recent
sediments from the Eastern Mediterranean Sea ____________________________39
5. Influence of diagenesis on sedimentary δ15N in the Arabian Sea over
the last 130 kyr ________________________________________________________59
6. Conclusions and outlook ________________________________________________83
Acknowledgements ________________________________________________________86
Figure captions ___________________________________________________________87
Table captions ____________________________________________________________91
List of abbreviations _______________________________________________________93
References _______________________________________________________________94
Curriculum Vitae _______________________________________________________ 110
v
1. Introduction
1.1. The marine nitrogen cycle
Reactive inorganic nitrogen, together with carbon and phosphorus, is the most important
nutrient for plant growth. In marine phytoplankton and in seawater these nutrients occur in the so
called Redfield-Ratio of 106 C : 16 N : 1 P, implying an equilibrium and that nutrient availability
in this proportion promotes optimal growth (Redfield, 1934). Whereas carbon is assimilated from
the abundant atmospheric pool and from CO2 dissolved in seawater, atmospheric dinitrogen (N2)
is bioavailable only for specialized prokaryotes containing the enzyme nitrogenase (diazotrophic
N2 fixation; e.g., Sprent and Sprent, 1990); for other primary producers it has to be present as
nitrate (NO3-), nitrite (NO2-), ammonium (NH4+), or dissolved organic nitrogen (DON) to be
assimilated. Hence, either phosphate or reactive N limit primary production in large parts of the
ocean (e.g., Tyrell, 1999), and only in specific areas production is limited by micronutrients such
as iron (e.g., Martin et al., 1990).
Whereas supply of phosphate to the ocean is governed by external inputs and internal
recycling (Benitez-Nelson, 2000), nitrate supply and turnover in the ocean is considerably more
complex. A schematic overview of the marine N-cycle is given in Figure 1.1 and briefly explained
in the following paragraph.
The major source of reactive N in marine systems is atmospheric N2 that is fixed as NH4+ and
further assimilated to organic N. Detrital organic N will either (i) settle through the water column
to the sediment, be buried and taken out of the marine N cycle, or (ii) be remineralized to NH4+
(ammonification) during organic matter breakdown in the water column or in the sediment. NH4+
is either oxidized (by nitrification) in several intermediate steps to NO3- (which is the most
important bioavailable N species and the largest oceanic pool of reactive N), or assimilated and
again transformed to organic N. Residual NO3- remains in the oxic water column/sediment, or is
reduced to NO2-(denitrification) under oxygen-deficient conditions in the water column or in the
sediment. NO2- is either (i) assimilated to organic N, or (ii) a reactant of the anammox reaction
and is transformed to N2, or (iii) is reduced to N2 via NO and N2O (denitrification). Gaseous
species such as N2O and N2 will be either fixed and assimilated, or released to the atmosphere and
leave the marine N-pool. Note that all transformations in the marine N-cycle are biologically
mediated processes. For an in-depth description of the intricacies of the natural N cycle, the reader
is referred to Canfield et al. (2005).
1
Chapter 1
Figure 1.1: The marine nitrogen cycle
On a global scale and over time scales of centuries to decades, natural sources and sinks of
marine bioavailable N species should be in equilibrium (e.g., Codispoti et al., 2001). Diazotrophic
N2 fixation and terrestrial/riverine inputs (the major sources) are balanced with N-loss during
denitrification and nitrification, and burial of particulate N in the sediment (e.g., Brandes and
Devol, 2002), although this balance has recently been questioned (e.g., Codispoti et al., 2001;
Codispoti, 2007). Over the last decades, anthropogenic input of fertilizers via rivers and
atmospheric total nitrogen inputs (NOx) became important contributors (Galloway et al., 2003;
Duce et al., 2008); they derive from industrial fixation of atmospheric N2 in the Haber-Bosch
process and fossil fuel burning that annually introduce 125*106 Tg of reactive N in addition to
natural sources (Gruber and Galloway, 2008).
Because the marine N-cycle determines the magnitude of the organic carbon pump in the
ocean, it is an important variable in validating and budgeting global variations of atmospheric
CO2 concentrations (Gruber and Galloway, 2008). Another link to global climate is given by the
powerful greenhouse gas properties of N2O (e.g., Badr and Probert, 1993).
2
Introduction
1.2. Stable N-isotopes (δ15N) - use as a proxy
14
The two stable isotopes of nitrogen,
atomic%
14
N and 0.3663 atomic%
15
N and
15
N occur in natural abundances of 99.6337
N in atmospheric N2. The ratio (R) of both isotopes is
expressed in relation to a standard as δ15N:
δ
15
(R
N (‰) =
sample
− R standard )
R standard
* 1000
R=
15
14
N
N
(1.1)
The standard is atmospheric N2, defined as δ15N = 0 ‰ (Mariotti et al., 1981).
As the
15
N atom is heavier than the
14
N atom and its vibrational frequencies are lower, its
molecular bonds are stronger and molecules containing
containing
14
15
N are usually less reactive than those
N. This leads to a slight discrimination against
15
N, in every non-equilibrium
transformation in the N-cycle (N2-fixation, NH4+/NO2-/NO3-assimilation, ammonification,
nitrification, denitrification, anammox). An exception of the
14
N preference is an inverse
fractionation occurring during the nitrite oxidation step (NO2- to NO3-) of nitrification (Casciotti,
2009). According to the “normal”
14
N preference, residual substrate of any reaction (e.g., NO3-
that is not assimilated) becomes enriched in 15N, whereas the product (assimilated N) is depleted
in 15N compared to the substrate. In a closed system the accumulated product and substrate will
have the same isotopic composition after complete turnover. This “Rayleigh”-type called
fractionation process is schematically illustrated in Figure 1.2, using the example of NO3assimilation in a closed system.
Figure 1.2: Rayleigh
fractionation
during
NO3- assimilation in a
closed system
3
Chapter 1
The degree of fractionation is process-specific and fractionation factors ε for several processes
in the N-cycle are listed in Table 1.1. ε is defined as (Rproduct/Rsubstrate – 1)*1000 and is
approximated by the difference in δ15N between the substrate and the initial instantaneous
product. The preference for
14
N or
15
N is expressed by negative values, if the lighter isotope is
preferred. Variability in ε is driven by physical parameters such as diffusivity and exchange rates
of pore waters, availability of the substrate, and biological factors related to enzymatic and “vital”
effects of organisms and assemblages of microorganisms involved.
Table 1.1: Fractionation factors ε for major transformation processes in the marine N-cycle. 14N preference is
expressed by negative values. ε -values for the anammox process have not been published yet; the published ε of
0‰ for ammonification (Kendall, 1998) appears unrealistic considering this study.
Process
Reaction
Fractionation
Reference
factor ε (‰)
N2 fixation
N2 NH4+
0 to -3.6
Ammonification
Norg NH4+
0
NO3- assimilation
NO3- Norg
-6 to -20
NH4+ assimilation
NH4+ Norg
-11 to -14
Carpenter et al., 1997
Brandes and Devol 2002
Kendall, 1998
Granger et al., 2004
Hoch et al., 1992
Voss et al., 1997
Nitrification
NH4+ NO2-
-14 to -38
NO2- NO3-
+12.8
Casciotti et al., 2003
(Ammonium oxidation)
Nitrification
Casciotti, 2009
(Nitrite oxidation)
Denitrification
NO3- N2
Anammox
NO2- + NH4+ N2
-22 to -30
Brandes et al., 1998
?
As a consequence, each N-pool has a specific isotopic signature (Table 1.2.) resulting from
the balance of source and sink terms and their isotopic fractionation.
4
Introduction
Table 1.2: Isotopic signatures of selected N-pools
Source
δ15N (‰)
Reference
Atmospheric N2 fixed N
0
Mariotti et al., 1981
Atmospheric NOx
-15 to 15
Kendall et al., 2007
Global ocean deep
water NO3-
5 ± 0.5
Sigman et al., 2000
Pristine rivers NO3-
-0.1 to 5
Mayer et al., 2002
Voss et al., 2006
Polluted rivers NO3-
5 to 22
Mayer et al., 2002
Voss et al., 2006
Johannsen et al., 2008
Fractionation during N transformation processes often modifies original source values. If
known, the δ15N values can be used as a proxy to trace N sources and turnover processes in core
records archiving past geological conditions, or to budget recent N-cycles. Systematic
relationships to other proxies can help to evaluate and quantify the impact of transformation
processes on δ15N.
Degradation of organic matter (OM) is a pathway of ammonification in a broader sense that is
usually accompanied by an enrichment in δ15N resulting from preferential loss of
15
N depleted
compounds (Altabet, 1996; Freudenthal et al., 2001; Gaye-Haake et al., 2005) during protein
hydrolysis (Bada et al., 1989; Silfer et al., 1992) and deamination (Macko and Estep, 1984). In
this study I investigate and define the relationship between OM degradation and 15N enrichment in
particulate N using paired data of δ15N and different indexes based on the amino acid composition
of the samples that are sensitive for OM degradation.
5
Chapter 1
1.3. The use of amino acid composition as a proxy for organic matter
preservation
Amino acids (AAs) are the building blocks of proteins and are ubiquitous in the OM
preserved in marine sediments. As they are differentially degraded during OM decay, changes in
the concentration of individual AAs and changes in the overall AA composition can be used as
indicators for the degree of OM degradation in younger sediments (Cowie and Hedges, 1992;
Dauwe and Middelburg, 1998). To gauge the OM degradation status from AA, molar
contributions of individual AA to total AA are expressed by two indexes that have been
established over the last ten years: (i) the Reactivity Index (RI; Jennerjahn and Ittekkot, 1997),
which is calculated as the ratio of aromatic to non-proteinogenic AAs (the latter do not exist in
living organic tissue but enrich during OM decay), and (ii) the Degradation Index (DI; Dauwe et
al., 1999) derived from a statistical treatment (principal component analysis) of data sets of 14
individual AAs. The DI will be introduced in Chapter 4 and both indexes will be introduced and
discussed in detail in Chapter 5.
As shown by the linear relationship between the DI and δ15N in surface sediments and sinking
particles from the Arabian Sea and from the South China Sea (Gaye-Haake et al., 2005; Gaye et
al., 2009), decomposition of AA goes in concert with an enrichment of δ15N in the residual OM
fraction. This systematic relationship will be used in the following to correct the δ15N record of
sediment cores for early diagenetic enrichment.
6
2. Working Areas, Projects and Thesis Outline
2.1. The Mediterranean Sea – a natural laboratory
The Mediterranean Sea is the smallest ocean - or the biggest enclosed marine basin in the
world. Relatively low exchange rates with the Atlantic Ocean, low concentration levels of nitrate
and phosphate, and the relatively fast turn over rates of limiting nutrients (as characteristic for
oligotrophic environments) make it an ideal place to study the impact of man-made and natural
environmental changes on external and internal nitrogen cycling on short time scales.
Atmospheric deposition of anthropogenic NOx increased by factor five to six since the 1920s
(Preunkert et al., 2003) and presently is considered to be the largest individual source of reactive
N that has changed the N inventory and N cycling of the Eastern Mediterranean Sea (EMS)
dramatically (Mara et al., 2009). The relative contribution of this new source to the reactive nitrate
pool in the mixed layer and the recycling rates of fixed N are calculated in Chapter 3 by budgeting
the δ15N of major involved N species.
For the past, repetitive deposition of organic rich sediment layers (sapropels) gives evidence
of dramatic changes in circulation patterns and nutrient cycling of the EMS (e.g., Emeis and
Weissert, 2009). It is widely accepted that Mediterranean sapropel formation is a result of water
column stratification and anoxic deep water conditions that developed as a consequence of
enormous fresh water inputs (e.g., Kullenberg, 1952; Cita et al., 1977). In most cases, these were
caused by northward shifts of the African monsoon belt (into the catchment area of the
Mediterranean Sea) that are driven by variations in solar radiation in the 23 kyr precession mode
of the Milankovitch cycles (Rossignol-Strick et al., 1982; 1983; Rossignol-Strick, 1985). To what
proportion enhanced productivity (as a consequence of enhanced nutrient supply; e.g., Calvert et
al., 1992) or preservation (due to the absence of oxygen in deep water; e.g., Cheddadi and
Rossignol-Strick, 1995) were responsible for OM accumulation is unclear. In the framework of
the European Science Foundation (ESF) project MERF (Marine Ecosystem Response to
Fertilization), my project attempted to develop and calibrate δ15N as a proxy for nitrogen cycling
and nitrate utilization in order to trace the impact of recent anthropogenic fertilization and to
explore temporal and spatial gradients in the possibly natural fertilized S1 sapropel (the most
recent) time slice (Chapter 4).
7
Chapter 2
2.2. The Arabian Sea – a key area of the global N-cycle
In contrast to the EMS, the Arabian Sea is one of the most productive areas in the world
ocean. Monsoon induced upwelling of nutrient-rich intermediate water to the euphotic zone
seasonally arises primary productivity up to more than 1.5 g C*m-2*d-1 (Barber et al., 2001). Due
to the remineralization of high export production flux in mid-water and oceanographic processes,
oxygen levels in subsurface waters and intermediate waters almost drop to zero resulting in a
permanent oxygen minimum zone (OMZ) roughly between 200 m and 1200 m water depth in the
central Arabian Sea (Wyrtki, 1971). As a consequence, large amounts of NO3- are reduced to N2O
and N2 by denitrifying and anammox bacteria and are no longer bioavailable in the marine Ncycle (Naqvi, 1994; Jaeschke et al., 2007). Accordingly, the Arabian Sea is both an important
source of the greenhouse gas N2O and an important sink of fixed marine N. Orbitally driven and
millennial-scale oscillations in monsoon intensity alter the OMZ in extent and intensity; this is
reflected for instance in sedimentary TOC and δ15N records (Altabet et al., 1995; Schulz et al.,
1998). My investigation aimed to appraise the diagenetic and environmental influences on δ15N in
basin-wide surface sediments and in core records of the last 130 kyr from various locations in the
Arabian Sea (Chapter 5).
The following individual chapters of this thesis constitute the basis of 3 articles that are under review
in, or submitted to peer-reviewed scientific journals:
Chapter 3
Emeis, K.-C., Mara, P., Schlarbaum, T., Möbius, J., Dähnke, K., Struck, U., Mihalopoulos, N. and
Krom, M.,: N-isotope ratios of nitrate, dissolved organic nitrogen and particulate nitrogen in
the Eastern Mediterranean Sea. Under review in Journal of Geophysical Research –
Biogeochemistry
My contribution to Chapter 3 (Emeis et al., under review):
-
recovery, analysis and interpretation of suspended matter samples and sediment trap
samples (MID 03)
8
-
determination of stable isotope pairs 15N/14N and 18O/16O in NO3 (part of the data set).
-
Co-work in discussion and manuscript preparation
Working areas, projects and thesis outline
Chapter 4
Möbius, J., Lahajnar, N. and Emeis, K.-C., 2010: Diagenetic control of nitrogen isotope ratios
in Holocene sapropels and recent sediments from the Eastern Mediterranean Sea.
Biogeosciences Discussions 7, 1131-1165.
Chapter 5
Möbius, J., Gaye, B., Lahajnar, N., Bahlmann, E. and Emeis, K.-C.,: Influence of diagenesis on
sedimentary δ15N in the Arabian Sea over the last 130 kyr. Submitted to Marine Geology
9
10
3. Isotope Ratios of Nitrate, Dissolved Reduced and Particulate Nitrogen in
the Eastern Mediterranean Sea Trace External N-inputs and Internal Ncycling
Abstract
The eastern Mediterranean Sea (EMS) is an unusually nutrient-poor ocean basin where the
15
N/14N isotope ratios in many compartments of reactive N are lower than in comparable oceanic
settings. To elucidate possible reasons, we determined stable isotope ratios in nitrate, suspended
particulate, total dissolved reduced nitrogen, and sinking particulate N for stations across the
Eastern Mediterranean Sea (EMS) occupied in January and February 2007. The absolute level of
δ15N integrated over all reactive N compartments in waters of the EMS is very low (grand average
2.6‰) compared to other oceanic environments. The δ15N of deep-water nitrate was lower than
nitrate generally found in the global ocean, whereas the δ15N of particulate N and reduced
dissolved N was significantly higher than that of nitrate and particulate N intercepted in sediment
traps. We infer that extensive mineralisation is the immediate cause of the isotopic makeup of
reactive N in deep water. Partial nitrate assimilation in the mixed layer had reduced
concentrations and raised δ15N and δ18O in nitrate, but an additional external source of nitrate with
a low δ15N, probably anthropogenic NOx rather than fixed nitrogen, was required to describe the
isotope distribution adequately. To explain the observed nitrate isotope anomaly ∆(15,18) in the
mixed layer, either the ammonium formed by OM breakdown must be predominantly nitrified, or
atmospheric NOx with characteristically high δ18O was present.
11
Chapter 3
3.1. Introduction
The modern Eastern Mediterranean Sea (EMS) is a highly oligotrophic oceanic environment
(Béthoux, 1989; Antoine et al., 1995) where primary production is approximately half the values
observed in the mid-ocean gyres such as the Sargasso Sea or the Northeast Pacific (Krom et al.,
2003). This low productivity is caused by the anti-estuarine circulation in the EMS: Modified
Atlantic Water (MAW) flows in through the Strait of Sicily at the surface, sinks in the eastern part
of the EMS and feeds the Levantine Intermediate Water (LIW), a high temperature, high salinity
water mass at depths between 200 and 500 m that leaves the EMS as a westward current through
the Strait of Sicily. This intermediate water mass collects and exports the bulk of mineralisation
products sinking out of the mixed layer. Below (>500 m) this surface circulation cell is the
Eastern Mediterranean Deep Water (EMDW) (Wüst, 1961; Malanotte-Rizzolli and Bergamasco,
1989) which forms when LIW entrained into surface waters of the two northern sub-basins
(Aegean or Adriatic Seas) gains sufficient density after winter cooling (Lascaratos et al., 1999).
The export of nutrients with LIW causes the deep waters of the EMS to be nutrient depleted
relative to the deep water in all other parts of the global ocean: Nitrate concentrations in EMDW are 46 µM compared to 8 µM in deep water of the adjacent Western Mediterranean Sea, and 20 µM at
similar depths in the Atlantic Ocean. The EMS is also unusual in that it has a high nitrate:phosphate
ratio (28:1) in the deep water (Krom et al., 1991) and is phosphate-starved (Krom et al., 2005a). As a
result, primary production during the winter phytoplankton bloom is limited by phosphate rather than
by nitrate (Krom et al., 1991). In winter, the phosphate in the surface water is entirely depleted while
excess nitrate remains (Kress and Herut, 2001). However, soon after the seasonal thermocline
develops in spring, both nitrate and phosphate become depleted in the photic zone (Kress and Herut,
2001; Krom et al., 2005b) and phytoplankton productivity becomes N- and P- co-limited (Thingstad et
al., 2005; Zohary et al., 2005).
Another unusual feature of the EMS is the low δ15N levels of deep-water nitrate, suspended
matter, and surface sediments in the EMS (Struck et al., 2001; Pantoja et al., 2002; Coban-Yildiz
et al., 2006) relative to values in other open-ocean environments. δ15N expresses the ratio in the
abundance of 15N and 14N in relation to that ratio (R) in a standard: δ15Nsample = (Rsample/Rstandard1)*1000 in ‰; the international standard being δ15N air N2 = 0‰. The low levels of δ15N in the
EMS are unusual, and three possible reasons have been suggested. The first is incomplete nitrate
utilisation (Altabet and Francois, 1994; Struck et al., 2001): Because most biological processes
discriminate against
15
N, incomplete nitrate uptake during phosphate-limited assimilation by
phytoplankton would result in products (such as particulate nitrogen, PN, or dissolved organic
12
Isotope ratios of nitrate, dissolved reduced and particulate N in the EMS
nitrogen, DON) depleted in 15N, while unprocessed residual nitrate would be enriched in 15N over
the original nitrate. Together, residue and products have the original nitrate signature, and
establishing isotope fingerprints of all compartments is a tool to quantify individual processes in
the N-cycle (Altabet, 1988). Establishing the relationships between mixed layer nitrate levels and
its δ15NO3- helps to decide if supply by deep winter mixing and subsequent nitrate utilisation
governs isotopic compositions entirely, or alternatively to identify situations where other sources
besides thermocline nitrate contribute to nitrate in the mixed layer.
One of the possible external sources that has been invoked to explain low δ15N levels (and high
N:P ratios) in the EMS is significant levels of diazotrophic N2 fixation (Sachs and Repeta, 1999;
Pantoja et al., 2002), because newly fixed nitrogen is isotopically depleted (δ15N = -2 to 0‰)
(Minagawa and Wada, 1986; Montoya et al., 2002). This interpretation is in line with other authors´
inference from N:P ratios that N2 fixation is the most important source of new nitrogen to the EMS
(Béthoux and Copin-Montegut, 1986; Gruber and Sarmiento, 1997; Mahaffey et al., 2005). At present,
however, observational evidence for N2-fixation in the EMS is ambiguous (Rees et al., 2006; BermanFrank et al., 2007; Ibello et al., in press), and annual input estimates range from nil (Krom et al., 2004)
to 120*109 mol/a (Béthoux and Copin-Montegut, 1986), the latter estimate being ~⅔ of the total
nitrate exported through the Straits of Sicily.
A third possible reason for low δ15N levels is significant atmospheric NOx input to the EMS that is
bordered by heavily industrialised regions to the North and is a busy shipping route. This basin is
distinct from other oligotrophic settings of the world ocean (Duce et al., 2008) in that atmospheric
deposition of NOx presently dominates the external nitrogen inputs (Krom et al., 2004; Mara et al.,
2009). Furthermore, regional atmospheric loads have increased dramatically over the last decades
(Kouvarakis et al., 2001; Preunkert et al., 2003; Fagerli et al., 2007), and the input of new nitrate from
the atmosphere to the surface water mass of the EMS presently accounts for almost 60% of reactive N
inputs (Krom et al., 2004; Mara et al., 2009). The atmospheric source (NO3- in dry and wet deposition)
in the EMS has a low annually averaged (flux-weighted) δ15NO3- of -3.1‰ (Mara et al., 2009) and
thus is a strong candidate for depressing the δ15NO3- levels in the EMS. Both N2 fixation and
atmospheric inputs have similarly low δ15N signatures and cannot be separated based on δ15N. But
whereas fixed N has to be oxidised to nitrate in the mixed layer, atmospheric NOx is a preformed input
of nitrate, and is known to have a high δN18O3- (Kendall, 1998; Sigman et al., 2009; Wankel et al.,
2009).
13
Chapter 3
The denitrifier method (Sigman et al., 2001; Casciotti et al., 2002) to simultaneously determine the
δN18O3- and δ15NO3-, has provided the analytical tool to investigate global (Sigman et al., 2009) and
regional (Lehmann et al., 2005; Sigman et al., 2005; Wankel et al., 2006; Casciotti et al., 2008; Knapp
et al., 2008) N-cycles that involve internal oceanic sources and sinks. The largest source of nitrate to
the oceanic mixed layer is thermocline nitrate. But in oligotrophic oceanic settings, nitrate regenerated
from particulate nitrogen (PN) and possibly from dissolved organic nitrogen (DON) via ammonia and
nitrite oxidation contributes significantly to primary production (Yool et al., 2007) and is thought to be
particularly important in the oligotrophic Mediterranean Sea (Diaz and Raimbault, 2000). Nitrification
imparts a characteristic isotopic signature on the δN18O3- and δ15NO3- of regenerated nitrate (Casciotti
et al., 2003; Wankel et al., 2006) that can be used to quantify the contribution of regenerated nitrate
(Sigman et al., 2009). Regenerated nitrate can either derive from PN or DON (Bronk, 2002; Knapp et
al., 2008). Should dissolved total reduced nitrogen (TRN, dissolved organic nitrogen and ammonia)
indeed be a key component in the recycling of reactive N in the EMS (Thingstad et al., 2005), its
isotopic composition must reflect active participation in the N-cycle (Knapp et al., 2005).
In this study, we determined the stable isotope ratios in nitrate (δ15NO3-, δN18O3-), in dissolved
total reduced nitrogen (δ15TRN, composed of DON+NH4+) and in particulate N (δ15PN) of both
suspended particles obtained from filtration from the mixed-layer and deep-water pools for a set of 17
stations occupied in the EMS in 2007 (Figure 3.1). Furthermore, mass flux and δ15N data of sinking
particles intercepted by sediment trap moorings in several periods from 1999 to 2007 at one of the
stations in the Ierapetra Gyre south of Crete are reported and discussed. Aims of this study are 1) to
provide a comprehensive data set of δ15N in several relevant N-pools of the water column, covering
pelagic and more nearshore provinces of the EMS and different water masses; 2) to explain the
processes which cause the unusual isotopic ratios in the deep water, and 3) to test if the unusually low
levels of δ15N are a consequence of incomplete nitrate utilisation in the euphotic zone of the EMS due
to P-limited phytoplankton blooms, or if internal N-recycling or external inputs (N2 fixation or NOx
inputs) are a significant source of mixed-layer nitrate.
14
Isotope ratios of nitrate, dissolved reduced and particulate N in the EMS
3.2. Materials and methods
Samples were obtained during an expedition with r/v METEOR in 2007, which visited 17 stations
in the EMS in January/February 2007 (Figure 3.1).
Figure 3.1: Map of stations occupied in the Eastern Mediterranean Sea during r/v Meteor cruise 71-3 (JanuaryFebruary, 2007). NIS stations referred to in the text are stations H07 to H12. The black square SE of Crete (water
columns sampling station Ier1) marks the location of the sediment trap deployment station (MID).
Water samples were taken with a rosette sampler equipped with a Seabird 911 CTD; an aliquot of
the bottom water at the sediment-water interface was also taken from multicorer deployments. Water
samples were immediately filtered through pre-combusted, rinsed GF/C filters and transferred into PE
bottles for shore based analyses of nutrient concentrations, total dissolved nitrogen (TDN) content and
N-isotopic signature (δ15TDN), and δ15N/δ18O of nitrate (δ15NO3- and δN18O3-). Samples for nitrate
isotope analysis were frozen on board ship for further analysis onshore, samples for TDN analysis
were oxidised immediately after filtration on the ship with persulfate and then stored frozen in brown
glass bottles until further analysis in the shore-based laboratory (Schlarbaum et al., in revision).
Samples for nutrient analyses were poisoned with 3.5% mercury chloride solution and stored at room
temperature. Nutrient and TDN concentrations were analysed immediately after the expedition using a
Bran+Luebbe Autoanalyzer 2 in the home laboratory and standard colorimetric techniques (Grasshoff
et al., 1999).
15
Chapter 3
Large water samples (10-50 L) were filtered through pre-combusted and tared GF/F filters for
analyses of total nitrogen concentrations in suspended solids (particulate nitrogen, PN), as well as for
δ15N of PN (δ15PN). Filters were frozen on board, lyophilised in the home laboratory and weighed
before further analysis.
On January 29, 2007, a mooring system MID-03 (Mediterranean Ierapetra Deep) was
deployed in the Ierapetra Deep off Crete (34°26.63' N, 026°11.58' E, bottom depth 3620 m)
(Figure 3.1). The system consisted of one sediment trap McLane PARFLUX MARK 7G-21 at
1508 m water depth (MID-03 Shallow) and one Kiel Sediment Trap K/MT 234 at 2689 m water
depth (MID-03 Deep). The cups were filled with filtered (GF/F, combusted) sea water from the
respective depths. In addition, 35 g l-1 NaCl and 3.3 g l-1 HgCl2 were added in order to avoid
diffusion and bacterial decomposition during the deployment. Sampling period started on 30
January 2007 and ended on September 03, 2007. The particle flux was sampled at intervals of 12
days and the mooring was recovered in September 2007. Subsequent to trap recovery, trapped
materials were sieved into >1 and <1 mm fractions, the wet sample material was filtered onto preweighed nucleopore filters and dried at 40°C. The dry weights of the <1 mm fraction are used for
calculating the total fluxes, and the filter cake was homogenised with an agate mortar prior to
analysis. The same site has seen intermittent sediment trap deployments at 2700 m water depth in
1999 (30 January to 13 April, 1999) and in 2001/2002 (5 November 2001 to 1 April, 2002)
(Warnken, 2003), and we use data from these earlier deployments to complement data obtained
from the last trapping period.
The samples for determinations of δ15NO3- and δN18O3- and δ15TDN were thawed in the shorebased laboratory, and nitrate isotopic composition was determined with the denitrifier method (Sigman
et al., 2001; Casciotti et al., 2002; Dähnke et al., 2008). The untreated filtered water samples or
persulfate-digested TDN samples were injected into a suspension of Pseudomonas aureofaciens for
combined analysis of δ15N and δ18O, or Ps. chlororaphis for δ15N analysis of TDN only. The resulting
N2O gas was flushed by purging the sample vials with helium, concentrated and purified on a
GasBench II (ThermoFinnigan), and analyzed on a Delta Plus XP mass spectrometer
(ThermoFinnigan). To avoid concentration-dependent fractionation effects, sample size in deep and
thermocline water samples was adjusted to achieve a final gas amount of 10 nmol. Many mixed-layer
samples, where nitrate concentrations were low, yielded smaller N2O amounts. For each sample,
replicate measurements were performed, and an international standard (IAEA-N3) was measured with
each batch of samples; we used a δ15N value of 4.7‰ and a δ18O value of 25.6‰ for IAEA-N3
referenced to Standard Mean Ocean Water (δ18O SMOW=0‰) (Böhlke et al., 2003; Lehmann et al.,
2003; Sigman et al., 2005). The contribution of nitrite was always below one per cent and will
16
Isotope ratios of nitrate, dissolved reduced and particulate N in the EMS
therefore not be considered further in our calculations, because the effect on δ18O values is negligible
(Casciotti and McIlvin, 2007). The standard deviation for IAEA-N3 was better than 0.2‰ (n = 5) for
δ15NO3- and better than 0.4‰ for δN18O3-. For further quality assurance of the results, we used an
internal potassium nitrate standard that was measured with each batch of samples. The standard
deviation for the internal standard was within the same specification for both δ15N and δ18O as IAEAN3. The duplicate analyses suggest an overall average standard deviation for δ15NO3- of 0.2‰ and for
δN18O3- of 0.3‰. The standard deviations of duplicate analyses of δN18O3- increased to of 0.5‰ in
samples with low nitrate concentrations in the mixed layer and upper thermocline. The majority of 85
samples analysed from the mixed layer gave unacceptable standard deviations and are not reported.
For calculations of the δ15N of total reduced nitrogen (TRN, the sum of DON and ammonia)
mass balance calculations were made using the measured concentrations (c) of nitrate and TRN,
and δ15TDN values of the oxidised sample, the reagent blank and the δ15NO3- of the original (not
oxidised) sample (Knapp et al., 2005):
δ15TRN = δ15TDN * c(TDN)/c(TRN) – [δ15NO3- * c(NO3-) + δ15NBlank * c(Blank)]/c(TRN)
Because of low concentrations and error propagation, the calculated standard deviation of δ15TRN
analyses ranged from 0.2 to 2.0‰ with a mean value of 0.9‰, and the calculated mean standard
deviation for TRN concentration was 1.7 µM. Note that δ15TRN was not corrected for any
contribution by ammonium. However, the only published data on ammonium from the EMS
found values in the range of 40-80 nM with no major trend with depth (Krom et al., 2005a). In the
same profiles the DON concentration was ~2-4 µM in deeper water, similar to the values
measured in this study. It is thus likely that most of the isotopic signal is due to DON.
Total particulate nitrogen concentrations (PN) in suspended solids and in sinking material
captured by the sediment trap (SPN) were analysed after high-temperature flash combustion in a
Carlo Erba NA-2500 elemental analyzer at 1100°C (Rixen et al., 2000). δ15N values were
determined using a Finnigan MAT 252 gas isotope mass spectrometer coupled to an elemental
analyzer. Pure tank N2 calibrated against the reference standards International Atomic Energy
Agency (IAEA)-N-1 and IAEA-N-2 was used as a working standard. The within-lab standarddeviation was found to be <0.2‰ based on a set of replicate measurements of 6 sediment samples
(Bahlmann et al., 2009).
17
Chapter 3
3.3. Results
3.3.1. Concentrations of N-bearing compounds
The study took place in winter when the water column at most stations was well mixed with a
thermocline situated at depths of around 100m (stations H07 to H12 in the northern Ionian Basin,
in the following abbreviated NIS, maximum water depth 1688 m, and station Sk01 in the northern
Aegean Sea; see Figure 3.1) to 250 m (stations in the deep Ionian Basin, Ierapetra and Herodotus
Basin, termed pelagic stations in the following) water depth.
Fluorescence profiles (Figure 3.2a) show elevated fluorescence (uncalibrated chlorophyll
concentrations) in the surface mixed layer typical of the winter phytoplankton bloom observed in
the EMS (Krom et al., 2003); these mixed layers were <100 m in the NIS station set and at station
Sk01, and >200 m deep in pelagic stations (Figure 3.2b). Concentration versus depth plots of
nitrate, phosphate, TRN and PN for all stations shown in Figure 3.3 in each of the individual
station profiles showed that the thermocline (coincident with the base of the fluorescence
increase) in all cases coincided with the nitracline. Based on the criterion of nitrate concentrations,
we distinguished samples above the nitracline, in the nitracline, and deep-water samples below the
nitracline for each station set (Table 3.1).
Nitrate concentrations were low but measurable in the mixed layers of both the NIS and the
pelagic stations, while phosphate concentrations (Figures 3.3a,b and Table 3.1) were below the
effective detection limit of the nutrient procedures being used (estimated to 0.05 µmol*L-1) (Li et
al., 2008). TRN concentrations varied around averages of 1.6 µmol*L-1 in the mixed layer of NIS
and pelagic stations, and 3.3 µmol*L-1 in the northern Aegean (Figure 3.3c and Table 3.1). PN
concentrations in the mixed layer of the NIS were on average 0.5 µmol N*L-1 (0.4 µ mol N*L-1 in
the mixed layer of the pelagic stations, and 0.5 µ mol N*L-1 in the northern Aegean) (Figure 3.3d,
Table 3.1).
18
Isotope ratios of nitrate, dissolved reduced and particulate N in the EMS
Table 3.1: Comparison of averages in water above, in and below the nitracline for samples taken in stations of
the Northern Ionian Sea (6 stations, Table 3.1a), at pelagic stations (10 stations, Table 3.1b) and at one station in
the northern Aegean Sea (Table 3.1c) in the EMS in January/February 2007 during Meteor expedition 71-3.
19
Chapter 3
Figure 3.2: Profiles of fluorescence (a; arbitrary units), nitrate (b) and phosphate (c) concentrations in the upper
400 m at 2 stations representative of NIS (H07) and pelagic stations (Her03) show stratification between 80 and
230 m water depth and indicate the biologically active mixed layer. An ongoing phytoplankton bloom in the
northern Ionian Sea (at station H07) is sustained by nitrate and phosphate provided from ongoing regional
thermocline deepening, whereas station Her03 illustrates the mature and thick mixed layer with very low nutrient
concentrations at pelagic sites.
The intermediate water mass had average nitrate concentrations around 3.5 µmol*L-1 and
average phosphate concentrations between 0.10 and 0.12 µmol*L-1; the mean N:P in the
intermediate water was around 30. TRN concentrations were around 1.5 µmol N*L-1 in
intermediate waters at NIS and pelagic stations, and 3.6 µmol N*L-1 in Aegean station. The TRN
concentrations in the intermediate water mass depth interval scatter widely due to measurement
uncertainties at the low concentration levels measured. PN (not determined in the intermediate
water mass of NIS stations) averaged 0.2 µmol N*L-1 at both pelagic stations and the single
northern Aegean station.
20
Figure 3.3: Concentrations of nitrate (a), phosphate (b), TRN (c) and particulate N in total suspended solids (d) plotted against water depth for all stations.
Chapter 3
Concentrations of most dissolved and particulate constituents in samples below the nitracline
to total depth were more uniform than in the biologically active surface (Figure 3.3; Table 3.1):
Average nitrate (4.5 to 4.8 µmol*L-1) and phosphate (0.16 to 0.23 µmol*L-1) concentrations over
all station sets resulted in an average N:P ratio of >27, similar to previous measurements across
the EMS (Kress and Herut, 2001). Average TRN concentrations varied around 1.3 µmol N*L-1 in
deep water at NIS and pelagic stations, but were elevated (average 3.6 µmol N*L-1) in the
northern Aegean. Average PN concentration was below 0.2 µmol N*L-1 in all waters below the
nitracline to total depth.
The two sediment traps deployed over 216 days from February to September 2007 at station
Ierapetra (SE of Crete, Fig. 1) monitored a total SPN flux in the shallow trap (at 1500 m) of 5.7
mmol N m-2 (or 26.3 µmol N*m-2*d-1) over that period, while the deep trap (at 2700 m) collected
1.4 mmol N m-2 (6.5 µmol N*m-2*d-1) of SPN over the same period. Earlier sediment trap
deployments at the MID location (Warnken, 2003) reported similar N fluxes in the deep traps
(MID-01, February to April 1999, 4.5 µmol N*m-2*d-1; and MID-02, November 2001 to March
2002, 6.0 µmol N*m-2*d-1); unfortunately, sampling in the shallower trap failed during those
deployments.
3.3.2. Isotopic composition of N-bearing compounds
At all stations, the isotopic composition of N-bearing species differed not only between the
mixed layer and the intermediate and deep water masses (Figure 3.4, Table 3.1), but also among
station sets. Mixed-layer nitrate had average δ15NO3- of 5.6‰, 2.3‰, and 4.1‰ at NIS, pelagic,
and northern Aegean stations, respectively; δN18O3- was 10.7‰ at NIS stations, and 5.2‰ at
pelagic stations; no sample yielded a reliable value in the northern Aegean (Table 3.1).
Particulate N in the mixed layer was low (δ15PN average of all stations = 1.3±0.7‰), and TRN in
the surface layer (average δ15TRN of all stations = 0.0±1.1‰) was roughly similar to PN in
isotopic composition.
22
Figure 3.4: Isotopic composition δ15NO3- (a), δN18O3- (b), δ15TRN (c), and δ15PN (d) plotted against water depth for all stations.
Chapter 3
The average isotopic composition of nitrate in samples from the nitracline was similar in all
station sets: δ15NO3- averaged 1.6 and 2.1‰ in samples below the mixed layers of the pelagic and
NIS stations, respectively, and 2.0‰ in the thick nitracline in the northern Aegean. δN18O3- of
samples from the nitracline had averages of 5.1‰ (NIS stations), 5.0‰ (pelagic stations), and
4.4‰ in the northern Aegean nitracline. Values of δ15TRN and δ15PN were higher in the
thermocline samples than in the mixed layer (Table 3.1).
In deep waters, δ15NO3- averaged between 2.0‰ and 2.3‰ in the three stations sets, and
δN18O3- was on average between 3.2‰ (NIS stations) and 3.8‰ for pelagic sites, respectively,
and these deep water values were not statistically different. Suspended matter in deep waters at all
sites had high δ15PN averages of 7.2‰ to 8.1‰, while δ15TRN measurements averaged between
2.4‰ and 6.7‰ over all sites. We attribute part of the high variability to measurement artefacts at
the low concentrations encountered. Samples of deep waters below the nitracline at pelagic sites,
where a relatively large sample pool was measured, had an average δ15TRN of 6.7±3.5‰ (Table
3.1).
The δ15SPN of material captured by the two sediment traps during the deployment in 2007
(MID-03) differed significantly from δ15PN of suspended matter collected by filtration in deep
water, and was essentially the same as PN found in the mixed layer during the expedition: The
upper trap had an average δ15SPN of 0.9±0.8‰ within the 216 days sampling period (1.0‰ flux
weighted), whereas the lower trap average δ15SPN was 0.8±1.0‰. This is somewhat lower than
δ15SPN for the 1999 deployment (MID-01; δ15SPN =2.2±0.4‰), but is in agreement with data
from trap MID-02 (deployment 2001/2002) that collected sinking particles with a mean δ15SPN of
1.2±0.6‰ (Warnken, 2003). Together the three deployments cover a period from November to
October, and thus the biologically active season for which we can expect the bulk of particle
transport out of the mixed layer to occur. In a composite annual cycle constructed from all three
deployments at 2700 m water depth (Figure 3.5, Table 3.2), SPN- flux peaks in March, whereas
highest δ15SPN (1.8‰) are indicated for the months of November and December. In general,
seasonality in δ15SPN is subdued and varies around an annual average of 1.1‰ with low values
from April to May (0.3-0.5‰) and higher values (1.2–1.8‰) later in the year from June to
December.
24
Isotope ratios of nitrate, dissolved reduced and particulate N in the EMS
Figure 3.5: Composite seasonal diagram of SPN fluxes (squares) and δ15SPN (circles) and their standard
deviations at 2600 m water depth at station Ierapetra (34°26´N, 26°11´E, water depth 3750 m). Fluxes (squares,
black line and s.d. in in red) and δ15SPN (circles, grey line, s.d. in blue) for three deployment periods (MID-1:
01/30/1999 to 04/13/1999), MID-2: 11/05/2001 to 04/01/2002 and MID-3: 01/30/2007 to 09/05/2007) at 2700 m
water depth have been assembled in a surrogate annual cycle.
Table 3.2: Composite seasonal course of N-flux and δ15N of sinking particulate matter at station MID
(Ierapetra), water depth 2700 m. Data are from three separate and discontinuous deployment periods starting in
1999 and ending in 2007, as explained in the text.
25
Chapter 3
3.4. Discussion
Table 3.3 is a compilation of inventories and δ15N of different water masses in the EMS and in
different compartments of reactive N based on our data, according to which the mass-weighted
and depth-integrated δ15N of the EMS is 2.8‰. Judging from that value, the inventory of reactive
N in the EMS is fundamentally different from the inventories of other oceans or regional marine
systems studied so far, because the overall level of δ15N is clearly lower than elsewhere. In the
following, we will first compare our new measurements on the individual compartments of N in
the EMS with previous data and those of other areas, discuss possible reasons for the unusual 15N
depletion, and finally test these possible reasons against the isotopic evidence.
Table 3.3: Estimate of reactive N-inventories of the EMS in different depth intervals and mass-weighted δ15N of
the different components (nitrate, particulate nitrogen PN, and total reduced nitrogen, TRN). We calculated
inventories (given in gigamol N) based on interval water volumes for the EMS without Adriatic and Aegean subbasins (R. Grimm, pers. comm., 2009) and weight δ15N values by average concentrations found during our
cruise. The last column is the integrated and mass-weighted δ15N over all components of reactive N for each
interval, the last line are the integrated inventories and δ15N of the entire water column.
3.4.1. Isotopic composition of reactive nitrogen in sub-nitracline and deep water masses
The starting point for our discussion is the isotopic composition of the large reactive nitrogen
in the deep-water pool and the depth interval below the nitracline and in the deep-water pool,
which integrates the isotopic signal over all internal and external nitrate sources over the deepwater residence time of 50-80 years (Roether et al., 1996). Differing from the pool in surface
waters (see below), this deep nitrate pool is isotopically homogeneous in the EMS. There are three
previous data sets of widely differing δ15NO3- in deep and intermediate water masses: One group
(Sachs and Repeta, 1999) determined δ15NO3- to -0.7±0.1‰ in two samples of deep water (depths
not given) from the EMS, analysed by the ammonia diffusion method (Sigman et al., 1997).
Another group (Struck et al., 2001), also by the ammonia diffusion method, gave a mean of
26
Isotope ratios of nitrate, dissolved reduced and particulate N in the EMS
7.3±2.8‰ in 6 samples of waters from between 200 m and 400 m water depth. A third group
(Pantoja et al., 2002), who used the denitrifier method, show 6 data points below 500 m in the
EMS but give a δ15NO3- of 2.5±0.1‰ of only 4 samples (in the text of that publication, the mean
is given as 2.4‰).
Our data confirm that δ15NO3- of nitrate in the deep water of the EMS (average of samples
below 500 m) is indeed lower (2.2±0.3‰; n=68) compared to the narrow range of 4.7 to 5.4‰ for
global ocean deep-water nitrate (Sigman et al., 2009), and our data set further shows that nitrate in
the LIW water mass from the base of nitracline to 500 m is even more 15N-depleted (1.8±0.4‰;
n=29) (Table 3.3). Deep water nitrate in the western Mediterranean Sea has a δ15NO3- of 3.0
±0.1‰ below 1500 m water depth (Pantoja et al., 2002), while in the adjacent NE subtropical
Atlantic Ocean, δ15NO3- is around 5‰ in waters >800 m depth (Bourbonnais et al., 2009). The
residence time of deep-water nitrate in the EMS has been estimated to be 125 years (Mara et al.,
2009). The pool is not directly linked to deep waters outside the EMS. It is fed mainly by
mineralisation/nitrification of particles sinking from the mixed layer and by preformed nitrate
downwelled during deep-water formation without subsequent modification by denitrification. The
main nitrate sink is the LIW water mass that exports nitrate to the western Mediterranean Sea at
depths between 150 and 500m. Plausible external sources are N2-fixation, atmospheric NOx
deposition, and river runoff.
The comparatively small size of the deep-water nitrate pool in the EMS (5-6 µmol L-1), its
relatively short residence time compared to that of other oceans, and the low δ15N of external
nitrate sources (runoff and atmospheric deposition without N2 fixation have an annual loadweighted δ15NO3- estimated at 0.5 to 1.5‰) (Mara et al., 2009) mean that over one residence time
(~125 yrs) the deep-water δ15NO3- should be depressed to approach the δ15NO3- of inputs (Mara et
al., 2009). The only data of δ15NO3- available (also determined by the denitrifier method and with
the same internal standard deviation of the method) to gauge possible systematic time-dependent
changes are those for 4 samples taken in 1999 in the EMS (Pantoja et al., 2002). These data had
an average of 2.5±0.1‰ below 500 m depth. Average δ15NO3- in our samples taken from below
500 m water depth 8 years later is 2.2±0.3‰ (n=68). While the difference in isotopic ratio is in
the predicted direction, the magnitude of the change is not large enough to confirm the hypothesis.
We point out, however, that the interval-integrated and mass-weighted δ15N of reactive N in the
EMS (Table 3.3) has a gradient from lighter to heavier values with depth, and that the surface
layer, which has a much lower residence time than deep water (on the order of years), is within
27
Chapter 3
the δ15N range of estimated external N inputs. Furthermore the intermediate water which is being
exported from the EMS has a value (1.8±0.4‰) close to the external supply.
The EMS deep-water (>500 m) nitrate pool is further unusual in that is has higher δN18O3(3.7±0.9‰, n=25) than other deep-water pools in the world ocean (Sigman et al., 2009). In table 2
of that publication, available data for the δN18O3- of deep nitrate from a variety of other deep
oceanic environments are reported, and all are lower (1.8 to 2.8‰, except in the water depth
interval from 300 to 1500 m eastern tropical Pacific with a δN18O3- of 7.0‰) than the EMS deep
nitrate pool. One previous reported value for the δN18O3- of deep water in the westernmost
Mediterranean Sea (Sigman et al., 2009) was 2.6‰ for the depth range of 1500 m to the seafloor,
and 3.1±0.1‰ for a composite value of in parallel measurements of 4 samples in deep water.
A probable reason for higher δN18O3- in EMS deep water than in other deep ocean pools may
be that the EMS is a concentration basin where evaporation exceeds precipitation and river runoff
(E/P+R ratio of 1.2 in winter and 1.83 in summer) (Gat et al., 1996). The average δ18O in water
samples deeper than 500 m is 1.43±0.18‰ and 1.44±0.16‰ in the 0-500 m depth range (Pierre,
1999), and thus is around 1.4‰ heavier than in other ocean basins (LeGrande and Schmidt, 2006).
Based on experiments (Granger et al., 2004), it has been suggested that nitrification does not
involve dissolved O2 for the required electron transfer to oxidise NH4+ to NO2- and subsequently
to NO3-- (Sigman et al., 2009). These authors argue that it is very likely that the δN18O3- is
inherited from the δ18O of ambient water with a positive offset of approximately 2‰, and the
higher δN18O3- in the EMS is in agreement with that concept.
Concentrations of TRN (predominantly DON) in EMS deep and sub-nitracline intermediate
waters are also very low compared with other environments (Berman and Bronk, 2003) and match
low PN concentrations (Table 1). To our knowledge, no data exist on δ15TRN from deep waters of
the global ocean, but δ15TRN in the deep (>500m) EMS is clearly higher (6.0±3.7‰, n=39) than
in EMS surface waters. It is also higher than in the subtropical (3.9±0.4‰) and equatorial NW
Atlantic (4.1±0.6‰), and the subtropical NE Atlantic (2.6±0.4‰), but in the range of values
reported from shallower depths of the subtropical North Pacific (5.4±0.8‰) (Knapp et al., 2005;
Meador et al., 2007; Bourbonnais et al., 2009).
This 15N-enriched TRN in EMS deep water coexists with low concentrations of even more 15N
enriched suspended PN, whereas our sediment trap data suggest that the δ15N of material sinking
rapidly from the mixed layer is low over the entire year and – differing from other observations
28
Isotope ratios of nitrate, dissolved reduced and particulate N in the EMS
(Altabet, 1988; Gaye-Haake et al., 2005) – is apparently not enriched in
15
N during its passage
through the water column. There was a significant decrease in the flux measured at different water
depths, with the flux of particulate sinking N in the upper sediment trap being 5.7 mmol N m-2
over the period of 216 days, decreasing to 1.4 mmol N m-2 at 2700 m water depth (second trap), or
to roughly 25%. This implies a loss 75% of particulate N flux to disintegration and
remineralisation over a 1100 m water column. Because both the concentrations of PN and TRN in
ambient water are low, and also decrease with depth (or are at best invariant), most of the loss
must be to ammonification and rapid nitrification to nitrate. In consequence, both TRN and PN in
deep water are very likely enriched residues of mineralisation, from which some lighter product
originated. Because at the same time all meso-zooplankton size classes in EMS deep waters
(Koppelmann et al., 2009) and surface sediments (Struck et al., 2001) are also enriched, the
product with low δ15N is likely to be nitrate. The data illustrate the efficient cycling of labile PN
in this ultra-oligotrophic system resulting in depleted nitrate and enriched TRN and suspended
PN.
3.4.2. Nitrate in the surface layer: Testing the model of incomplete nitrate utilisation
The intermediate water nitrate pool provides the bulk of nitrate available for assimilation in
the euphotic zone of the EMS. The EMS is unusual in that the phytoplankton bloom takes place
over the winter period (October-March), as indicated in 2006/2007 by Chl-a concentrations from
satellites (http://reason.gsfc.nasa.gov/ OPS/Giovanni/ocean.aqua.shtml). Soon after the seasonal
thermocline breaks down in autumn (October), nutrients are mixed into the surface layer. Since
the winter period in the Mediterranean typically consists of short periods (few days) of cold and
often wet weather interspersed with several days of warm sunny weather, this results in an
immediate growth of phytoplankton which can be observed by remote sensing imagery and by
fluorescence and nitrate profiles of NIS stations (exemplified in Figure 3.2). After each mixing
event, nutrients are removed from the mixed layer, until all of the phosphate has been taken up
and excess nitrate remains, together with biomass and TRN produced; sinking particles exit the
mixed layer across the pycnocline. This incomplete assimilation of nitrate has been previously
seen as a mechanism to explain low δ15N values in sediments and suspended matter of the EMS
(Struck et al., 2001).
At all stations sampled during M71-3, there was excess nitrate remaining in the surface mixed
layer. The average nitrate remaining in the euphotic zone was between 0.24 and 0.48 µM, and the
actual amount of residual nitrate varied with location and stage of thermocline evolution. These
concentrations are similar to those found previously for the average winter residual nitrate across
29
Chapter 3
the Southern Levantine basin (0.6±0.5 µM) (Kress and Herut, 2001). In contrast, phosphate was
entirely depleted in most surface water samples, with all values being below the detection limit for
dissolved phosphate in samples from the EMS that have been preserved by freezing prior to
analysis (<20 nM) (Krom et al., 2005a). This depletion of phosphate in the surface mixed layer while excess nitrate remained - during the winter phytoplankton bloom has been used as direct
evidence that the waters of the EMS are P limited (Krom et al., 1991).
If the concept of incomplete nitrate utilisation is correct, it requires that residual nitrate and
products of assimilation together have the original isotopic mixture of nitrate provided by mixing
(Mariotti et al., 1981). Products of assimilation in the mixed layer are those parts of PN and TRN
that are present in excess of their concentrations below the thermocline (which were mixed up
during homogenisation), and SPN exported by sinking particles. If an initial mixed-layer nitrate
pool in a closed system is progressively assimilated, it will be enriched in
15
N (and
18
O) in the
course of assimilation, in analogy to the Rayleigh distillation process. The enrichment can be
approximated by (Mariotti et al., 1981):
δ15NO3-residual = δ15NO3-initial+ 15ε ×ln( f )
with f = [NO3-]residual/[NO3-]initial, and
15
ε expressing the fractionation factor (in ‰) between
product and substrate. Similarly, the isotopic composition of instantaneous and accumulated
products at any fraction of substrate remaining in the course of assimilation can be calculated
(Mariotti et al., 1981).
The range of
15
ε reported in the literature is large and differs for different primary
producers. A recent compilation (York et al., 2007) reports 15ε from -16 to 6‰ (negative values
meaning that 14N is prefentially assimilated), but is commonly assumed to be -5‰. Available field
and experimental data suggest equal separation factors
15
ε and
18
ε for nitrate assimilation
(Casciotti et al., 2002; Granger et al., 2004; Lehmann et al., 2005).
Simplifying the EMS mixed layer to closed systems and using the Rayleigh closed-system
approach, we can test if progressive assimilation of thermocline nitrate matches observed isotopic
enrichment in residual nitrate and the δ15N of instantaneous and accumulated products (calculated
according to Mariotti et al., 1981) for pelagic and NIS stations (Figure 3.6 and Table 3.4a, b). We
assume an initial nitrate concentration in the mixed layer of half the concentration below the
thermocline due to mixing of essentially nitrate-free surface waters with sub-thermocline waters,
and at the nitrate isotope signature of water below the thermocline (Table 3.1) before the onset of
phytoplankton assimilation. This assumed starting stock of nitrate is similar in value to previous
estimates for the amount of nitrate mixed into surface waters in the EMS (Krom et al., 2003).
30
Isotope ratios of nitrate, dissolved reduced and particulate N in the EMS
Figure 3.6: Plots of log-normal nitrate concentrations (X-axis) versus δ15NO3- (a) and δN18O3- (b) found in
samples in and above the nitraclines for individual station. The solid regression line in Figure 6a is for samples
in the northern Ionian Sea (NIS) where nitrate is replenished from thermocline deepening (r2=0.89). In these
samples, 15N in nitrate is enriched in the course of ongoing nitrate assimilation with an apparent separation factor
15
ε close to -3‰. In samples from the open Ionian Sea and Herodotus Basin (pelagic stations), where the mixed
layer was thick and nitrate poor at the time of sampling, ln(nitrate) and δ15NO3- are uncorrelated. In Figure 6b,
the correlation between ln(nitrate) and δN18O3- for NIS samples is also significant (r2=0.78), with an apparent
separation factor 18ε close to -5‰. There is no significant correlation in samples from pelagic stations and from
the northern Aegean station.
31
Chapter 3
Table 3.4: Calculation of the isotopic composition of residual nitrate (δ15N and δ18O) and instantaneous and
accumulated products of nitrate assimilation expected from Rayleigh-fractionation in closed systems (a = NIS
and b = pelagic mixed layers) with starting concentrations and isotopic compositions corresponding to
thermocline nitrate from Table 1. Note that the permil fractionation factors 15ε and 18ε have been set to -5‰.
Results of measurements are also shown.
a)
b)
Isotopic composition of substrate and products at NIS stations, where the thermocline was as
yet thin (<100 m) and deepening, was roughly balanced with respect to the δ15N of nitrate
supplied by thermocline mixing and products of assimilation found. However, the apparent
separation factor
15
ε suggested by the relationship between the natural logarithm of nitrate
concentration and δ15NO3- is only -3‰ (Figure 3.6a), whereas the apparent
18
ε is close to -5‰.
The Rayleigh model predicts that by the time that 73% of the initial nitrate in the NIS mixed layer
32
Isotope ratios of nitrate, dissolved reduced and particulate N in the EMS
is assimilated into PN, SPN and TRN (such as is the case in the average profile over all stations
there), the residual nitrate should have a δ15NO3- of 8.6‰, which is higher than the 5.6‰ found
(Table 3.4a), and a δN18O3- of 11.6‰ (found 10.7‰) The theoretical accumulated products of
nitrate assimilation (PN, TRN, SPN) should have a δ15N of -0.3‰. The mass-weighted δ15N
found in all products (including an estimate of the amount and δ15N of SPN exported from the
mixed layer based on the sediment trap data) was around 0.7‰. The found values are similar to
those predicted by a closed system approach for the NIS stations, although (as will be shown
below), the different values for the apparent separations factors
15
ε and
18
ε suggest that nitrate
isotopic composition can be better described by also including externally supplied N.
By contrast, the isotopic compositions in the pelagic stations that had a mature thermocline
and a thick nitrate-depleted mixed layer deviate significantly from a simple closed system: Only
14% of the initially mixed nitrate remained (Table 3.4b) and should have a δ15NO3- of 11.5‰, and
δN18O3- of 14.9‰, whereas we found 2.3‰, and 5.2‰, respectively. Expected accumulated and
instantaneous products should have a δ15N of 0.0‰ and 6.5‰, respectively. While the value
found for the mass-weighted δ15N of products (PN, TON, SPN: 1.0‰) could be a mixture of
accumulated and instantaneous products, the deviations in δ15NO3- of the nitrate-poor pelagic
mixed layer clearly require an additional source of nitrate with a low δ15NO3-.
3.4.3. Constraints from differences in δ15NO3- and δ18NO3At all stations (NIS and pelagic), a second and related indicator for a deviation from simple
enrichment due to assimilation of thermocline nitrate is the differential behaviour of δ15NO3- and
δN18O3- of residual nitrate in the mixed layer. This could be an effect of unequal separation factors
suggested by data for NIS stations (see Figure 3.6), but this assumption would run counter to
available evidence (Casciotti et al., 2002; Granger et al., 2004; Lehmann et al., 2005). The
deviation from the expected 1:1 enrichment line originating from the composition of nitrate at the
base of the thermocline in samples from the mixed layers may expressed as a nitrate isotope
deviation ∆(15,18) (Sigman et al., 2005) from the dual isotope composition of the nitrate source:
∆(15,18) = (δ15Nmeasured-δ15Nsource) -ε15/ε18 × (δ18Omeasured-δ18Osource)
Figure 3.7 is a depth plot of ∆(15,18) for different station sets in the EMS over the top 500 m
and is calculated by using the average δ15NO3- and δN18O3- of EMS deep water as δ15Nsource and
δ18Osource, respectively. Our data points are scarce in the nitrate-depleted mixed layers, but suggest
33
Chapter 3
an average ∆(15,18) of EMS mixed layer nitrate of around -3‰, and both station sets suggest a
decrease towards the sea surface. The negative values imply an additional source with lower
δ15NO3- than the residual thermocline nitrate, and suggest either nitrate generation via nitrification
while nitrate in the mixed layer is being assimilated, or external input of nitrate with a different
isotopic makeup than the nitrate being assimilated (Sigman et al., 2005).
Figure 3.7: Depth plot of ∆(15,18) of nitrate for different station sets in the upper 500 m of the water column.
3.4.4. Nitrification of fixed N or recycled N
Nitrification in the presence of nitrate assimilation, via ammonification of particulate N and/or
DON, is the standard mechanism to cause a negative ∆(15,18). In the few oligotrophic
environments studied so far, observed negative ∆(15,18) have been attributed to nitrification of
comparatively light PN produced from fixation and possibly DON (Casciotti et al., 2008; Knapp
et al., 2008; Bourbonnais et al., 2009). In these cases, the isotopic composition of (in this case
“new”) nitrate is set by the δ15N of fixed N (~-1‰) on the one hand, and by the δ18O of ambient
34
Isotope ratios of nitrate, dissolved reduced and particulate N in the EMS
seawater with a positive offset of 2‰, so that nitrate deriving from N2 fixation is added to the
mixed-layer nitrate pool with a δ15NO3- of ~-1‰ and δN18O3- of ~ 2‰ (Bourbonnais et al., 2009).
In the EMS with its higher δ18O of seawater, we would expect the recycled nitrate to have a
δ15NO3- equivalent to δ15SPN and a δN18O3- of ~ 3.4‰.
Partitioning of ammonium released from PN to either nitrification or ammonium uptake has
also been shown to cause a range of ∆(15,18), because both processes have different 15ε (Sigman
et al., 2005; Wankel et al., 2006). Negative ∆(15,18) in residual nitrate may ensue, because nitrate
returned from PN mineralisation via ammonium oxidation is relatively more depleted in 15N than
the residual nitrate pool, whereas its δ18O is pegged to ambient water. Because on the other hand
ammonium uptake has a lower preference for
15
N than nitrification, significant ratios of
ammonium assimilation versus nitrification will cause biomass and thus PN to be relatively
enriched over the recycled nitrate. Because we find no indication for enriched SPN or PN in the
mixed layers, and the sediment trap data suggest that SPN is on annual average even more
depleted in 15N than thermocline nitrate in the EMS, we believe that assimilation of mineralised
ammonium is unlikely to play a major role as a substrate for phytoplankton growth in the EMS
mixed layer, and as a cause for negative ∆(15,18).
3.4.5. Input of NOx
Aside from an internal source of nitrate, there must be a second source that potentially causes
negative ∆(15,18) in the mixed layer of the EMS, namely atmospheric inputs of NOx. Two
previous studies (Knapp et al., 2008; Bourbonnais et al., 2009) acknowledge (but dismiss) the
possibility that the negative ∆(15,18) observed in mixed layer nitrate of the oligotrophic Sargasso
Sea and subtropical NE Atlantic Ocean indicate atmospheric NOx inputs, which have very high
δN18O3- at low δ15NO3- (Kendall, 1998). Such an input, if it is not immediately assimilated due to
phosphate limitation, would introduce nitrate with low δ15NO3- and high δ15NO3- into the surface
mixed layer of the EMS. We have evidence which suggests that NOx inputs play a relatively
larger role in the EMS than elsewhere: Modelled average annual total NOx inputs in wet and dry
deposition to the EMS surface range between 200 and 400 mg N*m-2*a-1 with a pronounced N-S
gradient (http://www.emep.int). An estimated 14-29 mmol N*m-2*a-1 of industrial origin are thus
supplied annually to the EMS.
The atmospheric deposition on the island of Crete in winter months of 2006/2007 had a mean
δ15NO3- of -2.0‰ (Mara et al., 2009), and likely had a high δN18O3-: 13 samples of dry
35
Chapter 3
atmospheric deposition (no rain events recorded) collected on the Island of Crete during JuneSeptember 2007 had an average δN18O3- of 67.5+4.2‰ (Mara, unpubl. data). Although this data
set does not cover the time of predominantly wet deposition in winter, the values are within in the
range reported from other environments (Kendall, 1998) and from the nearby northern Red Sea
(Wankel et al., 2009). There, δ15N values of water-soluble nitrate in aerosol samples ranged from
−6.9‰ to +1.9‰ and δ18O was found to range from 65.1‰ to 84.9‰ with highest δ18O values in
the winter. The source of that nitrate were air masses deriving from the Mediterranean Sea and
western Europe (Wankel et al., 2009). Thus, although only <10% of nitrate in the mixed layer of
the EMS may have originated from the atmospheric source over 4 months, the effect on average
δN18O3- and ∆(15,18) is largely due to the small amount of nitrate in the mixed layer of the EMS
and the high δN18O3- of atmospheric inputs.
We used the isotopic signatures of internal (recycling) and external (N2 fixation and NOx)
inputs together with an estimate of N export flux in a conceptual steady-state mass and isotope
balance model to explore the effects of externally and internally supplied N on ∆(15,18). The
model is the same as that used in the subtropical SE Atlantic (Bourbonnais et al., 2009), and we
refer to that publication for details. We adapted the model to conditions in the EMS by choosing
appropriate boundary conditions for the thermocline nitrate isotopic composition (from Table 3.1)
and the N-export production in the EMS. We calculated hypothetical ∆(15,18) for a range of input
conditions from 0-40 mmol N from N2 fixation and 0-40 mmol N of recycled nitrate, both of
which acquire their δN18O3- from nitrification and thus have a δN18O3- of 3.4‰. Nitrate input
from N2-fixation needs to be 40 mmol*m-2*a-1 (>70% of the 56 mmol*m-2*a-1 PN export flux) to
create a ∆(15,18) of around –2‰; input of 40 mmol m-2 a-1 of recycled nitrate yielded a ∆(15,18)
of –3.6‰, which is close to the 3‰ observed. For an assumed NOx input of only 5 mmol*m-2,
which is well within the known atmospheric NOx flux, the model calculates a ∆(15,18) of –5.8‰.
A number of possible combinations of these external and internal sources can theoretically
result in the nitrate isotope anomaly observed in the EMS mixed layer. However, our preferred
interpretation of the data is that they represent a mixture of regenerated nitrate and NOx input,
because both are known to be inputs to the mixed layer in the necessary fluxes and isotopic ranges
to fully describe the changes observed. We consider significant input from N2 fixation unlikely for
the time of our expedition, both because we sampled a winter situation, and because N2 fixation
measurements across the region sampled here were very low when sampled in June 2007 (Ibello
et al., in press). There is ample evidence that primary production in the EMS is supported by
regenerated nutrients that are entrained into a microbial loop operating in surface waters (Zohary
36
Isotope ratios of nitrate, dissolved reduced and particulate N in the EMS
and Robarts, 1998; Thingstad et al., 2005). Nutrient budgets (Ribera d´Alcala et al., 2003) and
experimental work (Thingstad and Rassoulzadegan, 1999; Thingstad et al., 2005) both imply that
surface productivity is to a significant extent supported by regenerated nitrate. In addition, recent
genetic investigations suggest that ammonia-oxidizing Archaea in mesopelagic waters of the
Eastern EMS may have a central role in nitrification of ammonium liberated from particulate N
(De Corte et al., 2009). Our data are in agreement with the important role of recycled nitrate, but
we again stress that atmospheric NOx inputs to the EMS must be taken into account in mass-based
(Krom et al., 2004) and isotope-based (Mara et al., 2009) budgets of the N-cycle in the EMS.
3.4.6. Is TRN involved in N-cycling?
TRN is the largest pool or reactive N in the mixed layer and after nitrate the second largest in
the entire water column of the EMS (Table 3.3). Because ammonia levels are low in the EMS, the
major constituent of TRN in our samples is likely to be dissolved organic nitrogen, which is a byproduct of N2-fixation or of heterotrophy in the mixed layer (Berman and Bronk, 2003). It has
been discussed as a possible substrate for phytoplankton assimilation or ammonification and
nitrification processes (Bronk et al., 2007), and its isotopic composition should reflect TRN
cycling (Knapp et al., 2005). In a previous study from the Sargasso Sea, the small concentration
and δ15DON differences between the surface and subsurface waters suggested that DON there is
recalcitrant and the data indicated only limited DON turnover in that area (Knapp et al., 2005).
Although our data is somewhat limited, we see a trend of decreasing concentrations and
increasing δ15TRN between surface and deep-water samples that imply active participation of
TRN in reactive N turnover. Our interpretation of the similar trends in δ15N composition of PN
and TRN (except in the intermediate water mass, where our data are scarce; Tables 3.1 and 3.3)
suggest they are closely coupled, and that both reflect
15
N enrichment in the course of
mineralisation to produce 15N depleted nitrate in the deep water mass, and most likely even more
intensely in the intermediate water mass. There is supporting direct evidence of nitrification at the
nutricline based on observed nitrite peaks (Krom et al., 2005a).
The similarity of δ15PN and δ15TRN in the mixed layer is likely to reflect heterotrophy, where
DON is a by-product of PN recycling. The low δ15TRN found in the surface layer of the EMS is
in accord with rapid grazing and recycling of nutrients during the winter bloom. At pelagic sites,
the stable seasonal thermocline had developed prior to our expedition, and had caused the winter
phytoplankton bloom to cease, a condition typically reached at NIS sites in March/April. As a
result, phytoplankton in the surface layers of the EMS is N- and P-co-limited by May (Zohary et
al., 2005). According to current understanding, active grazing of small phytoplankton populations
37
Chapter 3
persisting on recycled nitrate continues in the mixed layer (Krom et al., 2005a; Thingstad et al.,
2005), with the result that over time the residual nitrate is converted into DON. This DON may
not directly be bioavailable to the phytoplankton, but is consumed largely by heterotrophic
bacteria (Bronk, 2002; Thingstad et al., 2005). No isotopic measurements have been made of PN
and DON in summer in the EMS, but we would predict that the PN and DON should be
isotopically enriched compared to values obtained in this study of a winter situation.
3.5. Conclusions
The total pool of reactive nitrogen (nitrate, dissolved total reduced nitrogen, and particulate
nitrogen) of the eastern Mediterranean Sea is unusually depleted in
15
N compared to the global
ocean. This must to a large extent be due to the isolation and anti-estuarine circulation of the EMS
that prevents communication with the global deep-water nitrate pool that is very homogeneous at
δ15NO3- ~ 5‰/ δN18O3- ~2‰ (Sigman et al., 2009). The low level of δ15N of the EMS reflects the
dominance of an isotopically depleted N-source, because processes that enrich
15
N in nitrate of
other oceans (such as mid-water denitrication) are not acting in the EMS. The low δ15NO3- in the
deep water contrasts with high δ15N in TRN (mainly DON) and suspended PN. It is inferred that
this is due to extensive and efficient mineralisation of the rapidly sinking fresh PN (sampled in
sediment traps) in the ultra-oligotrophic system. We hypothesize that the δ15NO3- in the deep
water may be decreasing with time in response to the depleted (mainly anthropogenic) external
supply of NOx, although the data set is too sparse as yet to be conclusive. The data in this study
were collected in the winter of 2006/2007, which coincided with a typical winter bloom at the
stations in the northern Ionian basin and with an (unusually early) mature thermocline at
maximum depth in the remaining pelagic stations. Using a Rayleigh-closed system approach, it
was concluded that it was not possible to explain the isotopic distribution of NO3-, TRN and PN
only by partial N uptake caused by the extant P-limited phytoplankton bloom. It requires in
addition a source of isotopically distinct nitrate, which may be internal (recycled nitrate) or
external (N2 fixation or NOx input). Possible ranges for the nitrate isotope anomaly ∆(15,18) in
the mixed layer calculated by a simple model point towards nitrification and/or a relatively small
(and realistic) contribution of atmospheric nitrate with characteristically low δ15N and high δ18O
as the most likely sources of additional nitrate. It is known that the EMS receives a significant
supply of anthropogenic NOx, which together with known recycling processes adequately
describes the isotope distribution. These patterns could also be caused by extensive N2 fixation,
but that would require high rates of diazotrophy in winter, when there are insignificantly low
levels measured in the region even in summer, and would be in conflict with our knowledge on
the amount and isotopic composition of anthropogenic NOx input to the EMS.
38
4. Diagenetic Control of Nitrogen Isotope Ratios in Holocene Sapropels and
Recent Sediments from the Eastern Mediterranean Sea
Abstract
The enhanced accumulation of organic matter in Eastern Mediterranean sapropels and their
unusually depleted δ15N values have been attributed to either enhanced nutrient availability which
led to elevated primary production and carbon sequestration or to enhanced organic matter
preservation under anoxic conditions. In order to evaluate these two hypothesis we have
determined Ba/Al ratios, amino acid composition, N and organic C concentrations and δ15N on
sinking particles, surface sediments, eight spatially distributed core records of the youngest
sapropel S1 (10-6 kyr) and older sapropels (S5, S6) from two locations. These data suggest that (i)
temporal and spatial variations in δ15N of sedimentary N are driven by different degrees of
diagenesis at different sites rather than by changes in N-sources or primary productivity and (ii)
that present day TOC export production would suffice to create a sapropel like S1 under
conditions of deep-water anoxia. This implies that both enhanced TOC accumulation and δ15N
depletion in sapropels were due to the absence of oxygen in deep waters. Thus preservation plays
a major role for the accumulation of organic-rich sediments enhanced primary production is not
needed for sapropel formation in the Mediterranean.
39
Chapter 4
4.1. Introduction
In the Eastern Mediterranean Sea (EMS) paleoceanographic record, the ratio of stable N
isotopes
15
N/14N (δ15N) of organic-rich sediment layers (sapropels) has often been used to
understand mechanisms and environmental conditions leading to past black shale formation in a
region which is at present an extreme nutrient desert. In a commonly accepted model, recurrent
Mediterranean sapropel formation is attributed to a stratified water column and anoxic deep water
conditions that developed as a consequence of enhanced fresh water inputs (Ryan, 1972; Cita et
al., 1977; Rossignol-Strick et al., 1982; Rohling, 1994; Emeis et al., 2000). The youngest sapropel
– termed S1 – formed between 9,800 and 5,700 years ago (de Lange et al., 2008) at water depths
below 400 m (Anastasakis and Stanley, 1986) with inferred deep-water anoxia everywhere below
1800 m water depth (de Lange et al., 2008). The sedimentary total organic carbon concentration
(%TOC) reaches more than 30% in some Pliocene sapropels (Emeis et al., 1996) and is
approximately 2% in S1 (Murat and Got, 2000). Present-day pelagic surface sediments have a
range of %TOC from 0.2- 0.6% only (see below). The higher TOC content (which is associated
with faunal, floral, and geochemical changes) of sapropels has been attributed to either greatly
enhanced surface productivity, which would suggest a different nutrient regime compared to
recent conditions (Calvert et al., 1992; Kemp et al., 1999; Mercone et al., 2001), or to a better
preservation of organic matter under oxygen-deficient bottom-water conditions at approximately
equal productivity rates (Cheddadi and Rossignol-Strick, 1995; Sachs and Repeta, 1999; Moodley
et al., 2005) or some combination of both. Based on stoichiometric calculations several authors
recently proposed that there is a causal link between anoxia and enhanced productivity via the
enhanced recycling of phosphate from sediments/and sinking particles in the water column under
anoxic conditions increasing N2- fixation (Tyrrell, 1999; Wallmann, 2003). This model combines
increased productivity during S1 deposition with enhanced preservation (Struck et al., 2001;
Slomp et al., 2002; Arnaboldi and Meyers, 2006; Emeis and Weissert, 2009).
A key argument for this hypothesis are the uncharacteristically low values of δ15N in all
sapropels, ranging from -5.1‰ to 2‰, whereas surrounding hemipelagic sediments that are poor
in organic carbon always have δ15N >4‰ (Milder and Montoya, 1999; Struck et al., 2001; Meyers
and Bernasconi, 2005; Arnaboldi and Meyers, 2006). The δ15N in non-sapropel sediments thus are
similar to pelagic sediments world wide that range from 5 to 15‰ (Holmes et al., 1996; Holmes et
al., 1997; Gaye-Haake et al., 2005). These δ15N of sediments integrate the isotopic composition of
the assimilated nitrate source, biological fractionation, foodweb dynamics, and isotope
fractionation during sediment diagenesis (McClelland and Valiela, 1998; Voss et al., 2005;
Altabet, 2007; Fry, 2007; Dähnke et al., 2008). As a paleoceanographic tool, δ15N is widely used
40
δ15N in Mediterranean sediments
to reconstruct the reactive nitrogen regime of the oceans over geological time scales (Altabet,
2007; Farrell et al., 1995; Ganeshram et al., 2002; Garvin et al., 2009; Jenkyns et al., 2001;
Kuypers et al., 2004).
The fact that δ15N in sapropels is lower than in the hemipelagic carbonates has originally been
attributed to incomplete assimilation of excess nitrate from enhanced reactive N supply to the
euphotic zone by either river discharge or upwelling
(Calvert et al., 1992). Other authors
attributed massive addition of fixed nitrogen to intense P-recycling from anoxic sediments (Struck
et al., 2001), or to a compensation for nitrate loss by denitrification at suboxic interfaces between
surface and deep waters (Arnaboldi and Meyers, 2006). Interestingly, present-day surface
sediments, suspended matter, sinking particles and deep-water nitrate in the EMS all have light
δ15N that are similar to the δ15N of sapropel S1. Assuming that this indicates significant N2fixation in the modern EMS, (Sachs and Repeta, 1999) proposed that the S1-situation was similar
to the present-day situation, and that the low δ15N in S1 is due to the preservation of the original
δ15N signature of sedimenting newly fixed N. Preservation may play an important role because
sedimentary δ15N is known to become enriched during OM degradation resulting from
preferential loss of 15N depleted compounds (Altabet, 1996; Freudenthal et al., 2001; Gaye-Haake
et al., 2005). Although exact mechanisms remain unknown, kinetic isotope fractionation during
protein hydrolysis (Bada et al., 1989; Silfer et al., 1992) and deamination (Macko and Estep,
1984) very likely cause this enrichment.
The recent EMS nutrient and productivity regime appears to be an improbable setting for
black shale (aka sapropel) deposition, even if preservation of OM was enhanced by anoxia. The
EMS presently has a highly oligotrophic nutrient regime where the carbon export flux (6-12 g
C*m-2*y-1; Bethoux, 1989) is half of that of the Sargasso Sea. However, both modes of the EMS
(the modern nutrient desert and the sapropel mode) have the strikingly low δ15N values in
common: As shown in Figure 4.1, surface sediments in the EMS presently have δ15N-values
between 5‰ and 3.5‰ with an eastward decreasing gradient. This is explained by either
preferential assimilation of
14
NO3- (Struck et al., 2001) caused by an excess of nitrate over
phosphate in surface waters of the EMS (Krom et al., 2005), or by an eastward increasing supply
of fixed N (Pantoja et al., 2002). Although until recently fixed N2 was the strongest candidate to
explain the low modern δ15N levels, the observational evidence for significant diazotrophic N2
fixation is scarce and ambiguous (Mara et al., 2009). New data on the δ15N of NOx input from the
atmosphere suggest that the present-day N-cycle of the EMS may recently have been
anthropogenically changed: δ15N in nitrate in both dry and wet atmospheric deposition have
consistently negative δ15N compared to air N2, implying a strongly
15
N-depleted atmospheric
41
Chapter 4
source that contributes >50% of external nitrate inputs with a weighted annual δ15N of -3.1‰
(Mara et al., 2009). These authors propose that assimilation of the
15
N-depleted atmospheric
nitrate in the surface mixed layer and subsequent particle flux and mineralisation in the deep
water over the last 40-50 years accounts for the unusually low δ15N ratios found in deep-water
NO3- without the need of any significant N2 fixation. The nitrogen cycle in the EMS of today,
thus, is not likely to be in a natural state, and the similarity of modern and S1 δ15Ν values is due to
the fact that the dominant anthropogenic source today (δ15N = -3.1‰) and the postulated N2fixation source during S1 (-2 to 1‰; Minagawa and Wada, 1986) produce similar
15
N depleted
sedimentary δ15N values.
But what could explain the low δ15Ν values in S1 sediments with regard to input and cycling
of reactive N during sapropel deposition? It is essential that the signal encoded in δ15Ν is not
affected by post-depositional alterations, or that the diagenetic influence is known and can be
corrected for. Determinations of pristine δ15N, not affected by early diagenetic enrichment, have
been achieved by analyzing δ15N in diatom frustules (Sigman et al., 1999), foraminifer shells (Ren
et al., 2009) and biomarkers (Sachs and Repeta, 1999). In the case of the EMS and S1, data shown
in Moodley et al. (2005) already suggested a strong post-depositional enrichment in δ15N (by 45‰) associated with continued and progressive re-oxygenation of the uppermost section of the S1
sapropel (“burndown”). This appears to be a pervasive feature in oxic water columns (Saino and
Hattori, 1980; Altabet, 1988; Sachs and Repeta, 1999), implying that diagenesis under conditions
of variable bottom-water oxygenation would most seriously compromise the reliability of δ15N as
a proxy in environments with low sedimentation rates such as the Mediterranean Sea (Jung et al.,
1997).
The aim of this study is to quantify the mechanisms influencing δ15N during S1 sapropel
deposition and to identify post-depositional alteration. This is possible by using paired data of
δ15N, biological productivity (Ba/Al-ratios) (Dymond et al., 1992; Francois et al., 1995),
concentrations of bulk organic carbon and total nitrogen, as well as an index of organic matter
preservation based on amino acid composition. To decide if the isotopic signature is more related
to diagenesis/preservation or indeed reflects the nutrient regime at the sea surface (including
sources, availability and utilization of nitrate), we examined the S1 sapropel time slice at six
locations in different basins and at different water depths of the EMS. Data from sinking material
intercepted by sediment traps and surface sediments of the EMS are also examined for the impact
of early diagenetic alterations on δ15N during particle sinking and at the sediment surface. These
data and additional data from the Pleistocene S5 and S6 sapropels enable us to distinguish
between time slice specific and general tendencies in the behavior of δ15N. Specifically, they
42
δ15N in Mediterranean sediments
permit us to answer the following questions: Is δ15N in sapropels closely linked to preservation?
Were there spatial gradients in δ15N during S1 sapropel times, and were these gradients the same
as today?
4.2. Methods
Sinking particulate nitrogen (SPN) has been collected at the MID-station (Mediterranean
Ierapetra Deep) by sediment traps during three periods covering three to seven months from early
1999 to summer 2007. The traps used in this study were Mc Lane Mark VII and Kiel K/MT 234
type with a collection area of 0.5 m². Sampling intervals varied between 3.5 and 14 days, and the
traps were deployed in water depths between 1508 m and 2720 m. Detailed information about trap
position, depth and sampling intervals is given in Table 4.1.
Figure 4.1: Sampling sites in the Eastern Mediterranean Sea. Large dots: Sediment cores; Triangle: Ierapetra
Deep sediment traps and multicores; small dots: surface sediment samples. The color shading is δ15N (‰) in
surface sediments based on this study and Struck et al. (2001). The map is produced by using Ocean Data View
(Schlitzer, 2009).
Sediment samples were taken during Meteor cruises M 40/4, M 44/4, M 51/3 as well as RV
Minerva AD91 cruise (Giunta et al., 2003) and RV Aegeo cruise (Triantaphyllou et al., 2009) as
multicores and gravity cores. ODP 971C is an Ocean Drilling Program core recovered during Leg
160. Further details are given in Table 4.2. Core locations and the sediment trap deployments are
shown in Figure 4.1, superimposed are contours of present-day surface sediment δ15N in ‰.
43
Chapter 4
Table 4.1: Sediment trap moorings in the Ierapetra Deep: Position, water depth, sampling depth, deployment
period and trap type.
Mooring
Longitude
(°E)
MID 01
26.1792
MID 02
26.1900
MID 03 SH 26.1930
MID 03 DP 26.1930
Latitude
(°N)
34.4317
34.4417
34.4438
34.4438
Water depth Sampling depth
(m)
(m)
3750
2720
3600
2560
3620
1508
3620
2689
Deployment period
dd.mm.yy
30.01.99 - 13.04.99
05.11.01 - 01.04.02
30.01.07 - 22.08.07
30.01.07 - 22.08.07
Trap type
Mark VII
Mark VII
Mark VII
Kiel K/MT 234
Table 4.2: Sediment sampling locations and sample types: MC = Multicore, GC = Gravity Core, ODP = Ocean
Drilling Program Core.
Cruise
M 51/3
M 51/3
M 51/3
M 51/3
M 40/4
M 40/4
M 44/4
M 44/4
RV Ageo
RV Minerva
ODP 160
Core
# 562
# 563
# 569
SL 152
SL 67
SL 71
MC 284
MC 285
NS 14
AD 91-17
ODP 971C
Core type
MC
MC
MC
GC
GC
GC
MC
MC
GC
GC
ODP
Long. (°E)
19.191
23.499
32.576
24.611
27.296
23.194
26.097
26.179
27.047
18.586
24.683
Lat. (°N)
32.774
33.718
33.452
40.087
34.814
34.811
34.415
34.432
36.725
40.870
33.717
Waterdepth (m)
1390
1881
1294
978
2158
2788
4263
3702
505
844
2141
4.2.1. Analytical methods: organic carbon and nitrogen
Total carbon, organic carbon and total nitrogen were measured in duplicate by a Carlo Erba
1500 CNS Analyser (Milan, Italy). The precision of this method is 0.01% for total carbon and
0.002% for nitrogen. Organic carbon was analysed after removal of CaCO3 by 1N hydrochloric
acid (three times) with a precision of 0.02%.
4.2.2. Stable nitrogen isotopic ratio (δ15N)
The ratio of the two stable isotopes of nitrogen (15N/14N) is expressed as δ15N after
determining the abundance of the two isotopes in samples (after combustion and reduction of NOx
to N2) by mass spectrometry:
δ15 N (‰) =
(R
sample
− R standard )
R standard
R=
* 1000
15
14
N
N
(4.1)
15
The standard is atmospheric N2, defined as δ N = 0‰.
44
δ15N in Mediterranean sediments
δ15N values were determined using a Finnigan MAT 252 gas isotope mass spectrometer after
high-temperature flash combustion in a Carlo Erba NA-2500 elemental analyzer at 1100°C. Pure
tank N2 calibrated against the reference standards IAEA-N-1 and IAEA-N-2 of the International
Atomic Energy Agency and a sediment standard was used as a working standard. δ15N is given as
the per mil deviation from the N-isotope composition of atmospheric N2. Analytical precision was
better than 0.1‰ based on replicate measurements of a reference standard. Duplicate
measurements of samples resulted in a mean standard deviation of 0.19‰.
4.2.3. Amino acids
Total hydrolysable amino acids were analysed with a Biochrom 30 Amino Acid Analyser
after hydrolysis of 3-40 mg of sediments with 6 N HCl for 22 hours at 110°C. After separation
with a cation exchange resin, the individual monomers were detected fluorometrically. Duplicate
analysis according to this method results in a relative error of 4% for total amino acids. Further
analytical details are given elsewhere (Jennerjahn and Ittekkot, 1999; Lahajnar et al., 2007) .
Ratios and indices derived from the monomeric distribution of amino acid are used to estimate
the state of organic matter decomposition. The degradation index DI statistically evaluates relative
abundances of the 14 most common proteinogenic amino acids (Dauwe et al., 1999). The data
matrix used for a principal component analysis was the amino acid composition of 28 samples,
representing a wide range of environmental settings and degradation states from living plankton to
Pleistocene sediments. First axis factor scores derived from this analysis were taken as the DI
(Dauwe et al., 1999). To apply this index to our datasets, the molar percentages of individual
amino acids were inserted as vari into the following formula:
DI =
 var i − AVG var i 
 * fac.coefi
 STD var i
∑
i
(4.2)
where the deviation of the molar percentage of each amino acid from an average value (AVG) is
multiplied by an individual coefficient (fac.coef). AVG, fac.coef and STD (standard deviation) are
those given in the original publication (Dauwe et al., 1999). The sum of all operations yields the
DI of a given sample. Typical DI values range from 2 in well preserved samples (e.g. fresh OM
and sapropels) to -1.5 in highly degraded sediments.
45
Chapter 4
4.2.4. X-ray fluorescence
Concentrations of aluminum and barium were analyzed on fused (600 mg of sample diluted
with 3600 mg of lithium tetraborate) discs in an automated x-ray fluorescence spectrometer
MagixPRO (Panalytical) that is equipped with a Rh-anode. Loss on ignition was determined at
110°C and at 1000°C. Concentrations of major and trace elements were calculated based on a
calibration against international standards. Ratios of Ba/Al were calculated as weight ratios of the
elements barium and aluminum in the samples.
4.3. Results
4.3.1. Sediment traps
Mean fluxes of TOC and TN in the sediment traps at Ierapetra mooring MID varied between
0.57 and 0.81 mg*m-2*d-1 TOC (0.06 to 0.08 mg*m-2*d-1 TN) in the deep traps and 3.57
mg*m-2*d-1 TOC in the shallow trap (0.37 mg*m-2*d-1 TN) and followed the decrease of total flux
(Table 4.3) during particle sinking in the water column. On the other hand δ15N, TOC% and DI
show no depth dependence. Mean values of δ15N in SPN are between 0.76 and 2.17‰ (Table 4.3)
whereas TOC contents and DI range from 2.50 to 2.76% and from -0.21 to 0.32, respectively.
Table 4.3: Sediment trap moorings in the Ierapetra Deep: Mean values of δ15N, DI, TOC and total flux.
Mooring
δ15N
DI
(‰)
MID 01
MID 02
MID 03 SH
MID 03 DP
2.17
1.19
0.90
0.76
-0.21
-0.01
0.32
0.21
TOC
Total flux
(%)
(mg*m-2*d-1)
2.67
2.76
2.50
2.52
24.64
44.38
178.54
44.27
4.3.2. Surface sediments
The δ15N values of surface sediments are between 3.5 and 5.0‰ (Figure 4.1, Table 4.4) and
thus are significantly enriched over the sinking material intercepted by the sediment traps.
However, the general level of δ15N is low in comparison to values reported for other surface
sediments overlain by oxygenated deep waters (Holmes et al., 1996; Holmes et al., 1997; GayeHaake et al., 2005). Low TOC contents (0.26 to 0.75%) and poor preservation reflected by DI
46
δ15N in Mediterranean sediments
values that range from -0.99 to 0.47 are characteristic for the hemipelagic, well oxygenated open
marine sediments of the EMS.
Table 4.4: δ15N, DI and TOC values of recent surface sediments as well as from S1, S5 and S6 sapropel interval
(1) = estimated values; 2 )= mean values of entire sapropel).
Core
# 562
# 563
# 569
SL 152
SL 67
SL 71
NS 14
AD 91-17
# 562
# 563
# 569
SL 152
SL 67
SL 71
NS 14
AD 91-17
SL 71
ODP 971C
SL 71
Time
equivalent
recent surface
sediments
S1 sapropel2)
S5 sapropel2)
S6 sapropel2)
δ15N (‰)
DI
TOC (%)
5.00
4.50
3.50
4.30
4.40
5.00
3.901)
4.301)
1.74
0.52
1.34
2.81
1.25
1.57
2.3
3.56
-1.01
-0.69
0.17
-0.47
-0.71
-0.48
-0.99
-0.79
-0.64
n.a.
n.a.
0.62
n.a.
0.70
0.24
n.a.
0.96
0.46
n.a.
2.33
2.43
2.00
0.26
0.40
0.36
0.75
0.44
0.45
n.a.
n.a.
1.74
1.94
1.64
1.37
n.a.
2.29
0.86
n.a.
7.11
4.47
2.8
4.3.3. Downcore variations
The insert in Figure 4.2 displays DI and δ15N in the upper 10 cm of two multicores retrieved
from the Ierapetra Deep close to the trap location. Both cores show a downward enrichment in
δ15N, which is most pronounced in the uppermost 1.5 cm (MC 284 from 2.3 to 4.8‰; MC 285
from 3.5 to 4.6‰) and which goes in concert with an increase in amino acid degradation.
Downcore plots of analytical results for multicore #569 from Eratosthenes Seamount in the
eastern Levantine Basin that includes the S1 are shown in Figure 4.3a. The %TOC ranges from
1.2 to 2.0% between 22 and 30 cm core depth and indicates the extant sapropel S1 (dark shaded).
Above S1, %TOC ranges from 0.18 to 0.37%, below S1 we measured 0.58 to 0.76% in the socalled protosapropel (McCoy, 1974; Anastasakis and Stanley, 1984). A high Ba/Al ratio (seen as
the strongest argument for elevated productivity during S1 formation) (Thomson et al., 1999),
outlines the original extent of the sapropel in a bell-shaped maximum centered at 22 cm. In the
interval from 13 - 22 cm core depth, divergence of Ba/Al and TOC curves marks the burndown
zone above the visible S1 (light shaded), where downward progression of oxygen eradicated the
47
Chapter 4
high TOC concentrations, but did not affect Ba/Al (Thomson et al., 1995). The DI curve follows
that of TOC contents and displays a well preserved sapropel (DI = 0.4 to 0.9); OM is highly
degraded above the sapropel and especially so in the burndown zone (-1.7 to -0.5). In the
protosapropel, organic matter is less preserved than in the sapropel directly above (DI = -0.2 to
-0.1). The δ15N pattern is a mirror image of %TOC and DI curves with values between 4 and 5‰
above the sapropel (an exception is the low value of 3.5‰ of surface sediment), relatively light
δ15N (0.6 to 2‰) in the sapropel, and intermediate values (2.6 to 3.3‰) in the proto-sapropel.
Cores #562 and #563 (Figure 3.3b and 3.3c) show more or less the same patterns in %TOC, δ15N
and Ba/Al as core #569 and confirm it as typical examples of sapropel S1.
Figure 4.2: δ15N vs. degradation index (DI) in cores MC 284/285 from Ierapetra Deep and mean values from
MID sediment traps. Insert: High resolution records of δ15N and DI of multicores MC 284 and MC 285.
Mean values of δ15N, DI, TOC for the S1 sapropel at the eight locations examined and for the
recent surface sediments at these sites are listed in Table 4.4 (see Table 4.2 for locations and water
depths). Average δ15N in the S1 range from 0.52‰ in core #563 to 3.56‰ in core AD 91-17. The
DI indicates best preservation of amino acids in core SL 71 south of Crete (DI 0.96) and poorest
preservation in SL 152 in the northern Aegean Sea (DI 0.24). %TOC in S1 ranges from an
48
δ15N in Mediterranean sediments
average of 0.96% in NS 14 (Aegean Sea) to 2.29% in SL 71. Unfortunately, DI and TOC data are
not available for S1 in core SL 67, and DI data are lacking for cores AD91-17 and #563. The S5
(deposited after 127 kyr) and S6 sapropels (deposited after 176 kyr) from cores SL 71 and ODP
971 C (Table 4.4) have highest organic carbon concentrations (2.8 to 7.1%) and best amino acid
preservation, whereas δ15N is most depleted in these two older sapropels (δ15N = 0.17 to -1.01‰).
4.4. Discussion
Although the modern N-cycle of the EMS may not be in a natural state due to anthropogenic
inputs of atmospheric nitrate that currently dominates external N-inputs (Mara et al., 2009), this
reactive N has a δ15N of approximately -3‰ (Mara et al., 2009) and thus is very similar to fixed N
(-2 to 1‰; Minagawa and Wada, 1986) that most likely has been the prevailing N source to the
EMS in pre-industrial times (Sachs and Repeta, 1999). This coincidence in the δ15N of the
dominant N source is fortuitous, but it permits us to track the effects of diagenesis from the water
column to the sediment record.
4.4.1. Alteration of amino acids and δ15N in the water column and in surface sediments
Enrichment of δ15N during sinking of particles through an oxygenated water with increasing
water depth and at oxygenated sea floors is well known from other studies. Surface sediments
commonly show an enrichment of 3- 4‰ over material in the oceanic mixed layer and sinking
particles (Altabet, 1996; Gaye-Haake et al., 2005). Increasing δ15N together with progressive
amino acid degradation in the two multicores from Ierapetra Deep (Figure 4.2) illustrate an
enrichment of sedimentary δ15N during early diagenesis in the uppermost cm (which is still
affected by sea water) as observed in the Eastern Atlantic (Freudenthal et al., 2001). A plot of
δ15N vs. DI (Figure 4.2) highlights the close relationship between δ15N and degradation state for
sinking material (mean values and ranges from Ierapetra Deep sediment trap long term moorings
MID 01 – 03) and sediments. The linear regression (r² = 0.85) for the entire data set is highly
significant and indicates an enrichment of 3 to 4‰ in δ15N during OM decomposition from the
water column to the sediment in the modern EMS. Pronounced differences in preservation and
relative OM content are not evident between shallow and deep traps, contrasting with differences
between traps and surface sediments. This implies that degradation during particle sinking in the
water column (few days or weeks) has a much smaller effect than degradation at the oxic seafloor,
where organic matter is exposed for decades and centuries due to low sedimentation and sealing
rates. The observed decrease in total flux (the deep trap captured only one quarter of the shallow
49
Chapter 4
trap) thus reflects disaggregation and disintegration of sinking material without compoundspecific fractionation.
4.4.2. The record of S1
The systematic variations in OM preservation and δ15N in the modern EMS help us to
interpret the δ15N record of core #569 (Figure 4.3a). A plot of DI versus δ15N for samples from
this core (Figure 4.4) again indicates a highly linear relationship between the δ15N and OM
preservation. The well preserved and isotopically depleted sapropel S1 (core interval from 22 to
31 cm) defines one end of a slope at good preservation state and low δ15N that ends with strongly
degraded and
15
N enriched (burndown and post sapropel) sediments of the upper core section.
Recent surface sediment (0- 1 cm) and the protosapropel are moderately preserved and
moderately enriched in δ15N. In this plot, average values from sediment trap material plot close to
protosapropel and sapropel samples. This implies that the δ15N of primary produced organic
matter probably was similar during sapropel deposition as today because data from sediment traps
show that relatively fresh OM sinking through the water column today has an isotopic signature
similar to that of sapropels. This has already been postulated by (Sachs and Repeta, 1999), but
possibly for the wrong reason: The δ15N today is determined by atmospheric nitrate inputs,
whereas the S1 inputs must have been N2 fixation.
What is most important is the fact that major temporal fluctuations in the core record appear
to be largely controlled by OM preservation (as indicated by the DI) and not by reactive N
sources:
15
N enrichment in sedimentary N unambiguously occurs as an enrichment of
15
N in
residual OM during progressive mineralization (ammonification) in the sediment as shown by the
insert in Figure 4.4. This matches experimental evidence of changes in δ15N in algal material
exposed to oxic degradation (Lehmann et al., 2002). The sapropel deposited during anoxic
conditions at the seafloor or in the water column below 1800 m water depth (de Lange et al.,
2008), when decay was inhibited, retained the δ15N of primary production. The protosapropel that
developed under suboxic conditions is slightly degraded and slightly enriched in δ15N over the
sapropel base level. The δ15N of the burn-down zone above the visible sapropel originally must
have been equally low as in the visible sapropel but became enriched during re-oxygenation of the
sediment. Interestingly, recent sinking material from sediment traps is somewhat more degraded
and enriched in
15
N in comparison to the material in the S1 sapropel. It may not reflect the
original δ15N of primary produced OM because the traps sampling at depths of 1508 to 2720 m
intercept material that has already been slightly degraded and thus isotopically enriched.
50
Figure 4.3: Downcore plots of total organic carbon (%TOC), barium to aluminum ratios, degradation index (DI) and δ15N in multicores #569 (a), #562 (b) and #563 (c). S1
sapropel intervals and the burn-down zone above S1 are marked.
Chapter 4
Figure 4.4: δ15N vs. DI in core #569 and mean values from MID sediment traps. Insert: δ15N vs. N (%) in core
#569. Residual N becomes isotopically enriched during remineralization. Surface sediments and protosapropel
are marked in both plots.
Can the data help to reconstruct the δ15N of primary produced OM in pre-industrial times, i.e.
between the S1 sapropel and the recent sediment surface that may be contaminated with NOx? In
core #569, the pre-industrial interval (from 1 to 13 cm) has δ15N values between 4 and 5‰.
According to the slope of δ15N versus DI (Figure 4.4) and extrapolating to well-preserved OM (DI
between 0.5 and 1.0), we reconstruct an original δ15N of 0 to 2‰ in fresh OM produced at that
time, even though the modern atmospheric source was not yet in operation. This requires that
either (1) nitrogen fixation was significant in the pre-industrial EMS or that (2) other isotopically
sources with low δ15N, such as terrestrial/riverine N, were the major inputs. We speculate that the
EMS has been N-limited since millennia, and N-fixation occurred regardless of sapropel stages or
“normal” stages until the 20th century. During the last decades, atmospheric inputs of industrial
NOx replaced N-fixation as an N-source in a smooth transition – and the EMS became P-limited
due to excess external N loads. Because N-fixation and NOx both result in δ15N depleted OM, the
transition is not visible in the core record. One could argue that downcore enrichment in
15
N in
several EMS core tops (MC 284, MC 285, #562, #563 and #569) reflects the transition to the
52
δ15N in Mediterranean sediments
more depleted source of reactive N. However, the simultaneous increase of amino acid
degradation strongly suggests that the enrichment must be due to diagenesis.
4.4.3. Spatial gradients during the S1 sapropel time slice
Comparing the mean values of δ15N in S1 sapropels at eight locations in the EMS shows that
δ15N is not spatially homogeneous during the S1 sapropel timeslice; this is in contrast with records
for S5 (Struck et al., 2001) that all originated from deep basins. Five cores from the Levantine
Basin and the Ionian Sea (M40-4 67, M40-4 71 and M51-3 562, 563 and 569) are relatively more
depleted in 15N compared to nearshore locations in the Adriatic and in the Aegean Sea (AD 91-17,
NS 14 and SL 152). In line with the hypotheses formulated above, we also attribute this gradient
to preservational effects: The northern Mediterranean sub-basins have been suboxic, but not
anoxic, during S1 deposition, as is indicated by benthic foraminiferal assemblages and
abundances (Kuhnt et al., 2008; Kuhnt et al., 2007). Somewhat higher δ15N, less well preserved
amino acids, and lower %TOC than in S1 from the deep basins probably result from stronger
degradation during particle settling through the suboxic water column and at the suboxic seafloor.
Average δ15N and DI values of samples from the S1 time slice (insert Figure 4.5) follow the same
linear relationship between OM preservation and nitrogen isotope composition as was found for
recent samples from Ierapetra Deep (MC 284, MC 285, MID 01 – 03) and the paleo record from
Eratosthenes Seamount (M51-3 #569). But even if the trend is highly linear, some influence of
different nitrate sources/utilization may be present: Extremely 15N depleted OM in cores from the
deep basins may have been contributed mainly by nitrogen fixing cyanobacteria, whereas
locations closer to land may have been more influenced by terrestrial/riverine OM inputs that are
somewhat less depleted in 15N (0 to 5‰, Mayer et al., 2002; Voss et al., 2006). Possibly due to the
dominant atmospheric source today, recent Eastern Mediterranean surface sediments display a
more or less homogeneous pattern in δ15N without pronounced gradients from open marine to
nearshore environments or from north to south (see Figure 4.1).
53
Chapter 4
Figure 4.5: δ15N vs. DI for S1, S5 and S6 sapropels, core #569 and all Eastern Mediterranean surface samples.
Mean values.
4.4.4. The S5 and S6 sapropel
The older sapropels S6 and in particular the S5 have even higher organic carbon contents,
better preservation of amino acids, lower δ15N values, and greater thickness than the S1 sapropel,
and δ15N and DI for these layers in cores ODP 971C and SL 71 extend the linear relationship
between preservation and δ15N (Figure 4.5). This implies a similar isotopic signature of primary
produced PN for all sapropels, and also suggests that the PN produced during S5 was
uncharacteristically depleted (< -1‰). As in surface sediments and S1, the differences between
the sapropels must be due to differences in organic matter preservation that may have been caused
by variable degrees of anoxia during deposition: The depth of the oxycline and hence the oxygen
exposure time of the sinking material has been variable for each sapropel interval as indicated by
biomarker records. For the S5 sapropel (which has best preservation and lowest δ15N), molecular
fossils of anaerobic, photolithotrophic green sulphur bacteria (Chlorobiacieae) evidence a
relatively shallow oxic/anoxic interface at 150 to 300 m water depth (Rohling et al., 2006),
whereas during S1 sapropel deposition the interface has been at ~1800 m (de Lange et al., 2008).
54
δ15N in Mediterranean sediments
4.4.5. Alternative explanations for 15N depletion in sapropels
The negative δ15N in S5 is unusual even if N2-fixation and/or incomplete nitrate utilization
occurred, and we speculate that another source of isotopically light nitrogen may have contributed
to sedimentary N: Recent studies in the Black Sea, which is permanently anoxic at depth and thus
may be a suitable analogue for the S5 situation in the EMS, revealed that appreciable amounts of
depleted particulate N must be produced by chemoautotrophic bacteria at the transition from
suboxic to anoxic waters (Fry et al., 1991; Coban-Yildiz et al., 2006). This production is sustained
by either assimilation of ammonia (Fry et al., 1991; Coban-Yildiz et al., 2006) or may derive from
chemoautotrophic fixation of N2 (McCarthy et al., 2007; Fuchsman et al., 2008) and N2O
(Westley et al., 2006) that has been released by denitrification or anammox. Water column
profiles of suspended organic δ15N (SPON) (Coban-Yildiz et al., 2006) show a slight enrichment
from ca. 4‰ in the upper mixed layer to up to 8‰ at the oxycline, as expected during early stages
of OM decomposition in an oxygenated water column. Below the oxycline in the Black Sea,
however, δ15N values shift dramatically to extremely depleted values (-8‰) at the top of the
anoxic water body. A simultaneous increase in the mass of total SPON implies that 15N-depleted
OM must be newly produced. This was attributed to newly produced OM by the biomass of
chemoautotrophic bacteria utilizing NH4+ as their dominant N source (Coban-Yildiz et al., 2006).
Estimated fractionation factors for bacterial ammonia assimilation are between 11 and 14‰
(Hoch et al., 1992; Voss et al., 1997), explaining the extremely low δ15N values of bacterial
biomass.
A second mechanism leading to negative δ15N in S5 may have been the fixation of
15
N
depleted N2 and N2O that was released by dentrification and anammox. An unusual isotopomeric
composition of N2O in the suboxic zone of the Black Sea suggested that there has been
consumption and hence assimilation of depleted N (Westley et al., 2006), and N2 fixation has been
observed in samples from the suboxic water layer during incubation experiments (McCarthy et al.,
2007).
Considering the recent Black Sea as an analog of the EMS in its sapropel mode (i.e. anoxic
deepwater, stratification, reduced surface water salinity, low 15N values, TOC-rich sediments), it
appears plausible that similar processes also occurred at the redox boundary in mid-water during
sapropel S5 formation. The extremely low
15
N values in the S5 sapropel and as low as -5‰ in
older Pliocene sapropels (Arnaboldi and Meyers, 2006) are difficult to explain by either N2fixation from cyanobacteria or by the preferential uptake of 14N–NO3- as a result of NO3- excess
alone, but instead suggest a highly fractionating N-pathway such as chemoautotrophic NH4+
assimilation.
55
Chapter 4
Interestingly, δ15N values of recent deep Black Sea sediments (2-3‰; Reschke, 1999) are in the
high range of Mediterranean sapropels (-5 to 3‰). The reason may lie in the fact that main
sources of bioavailable N in the Black Sea are terrestrial and riverine inputs (McCarthy et al.,
2007), whereas fixation of atmospheric N2 probably was the main N-source during times of
sapropel formation in the Mediterranean Sea.
4.4.6. Higher than modern productivity during S1 deposition?
There is a long-standing debate on whether TOC enrichment in Mediterranean sapropels may
be solely due to better preservation under anoxic deepwater conditions, or whether it requires
enhanced primary production (Howell and Thunell, 1992; Cheddadi and Rossignol-Strick, 1995).
When comparing reconstructed TOC accumulation rates in the S1 sapropel with TOC fluxes in
the sediment traps, we found that the modern flux rates would indeed suffice to produce a
sapropel under appropriate preservation conditions: Sediment traps recorded a TOC-flux of 207 to
294 mg*m-2*y-1 in the deeper traps, and 1300 mg*m-2*y-1 in the shallow trap. On the other hand,
the TOC accumulation rate for the S1 sapropel of core #569 is roughly 450 mg*m-2*y-1, which
agrees with accumulation rates calculated for a larger set of S1 by de Lange et al. (2008). Not only
is the accumulation rate similar, but also %TOC of sinking material is in the range known from
the S1 sapropel (see Table 4.3 and Table 4.4). The EMS was suboxic in waters deeper than 400 m
(Anastasakis and Stanley, 1986) and anoxic below 1800 m during S1 time (de Lange et al., 2008).
Assuming a distinctly reduced or even inhibited decay of OM under anoxic conditions, and
considering that OM in the traps has been already degraded during passage of the oxic water
column, present-day export production in the EMS would suffice to create an S1 sapropel of the
correct δ15N.
Enhanced primary productivity during S1 has mainly been inferred from elevated Ba/Al ratios
(Thomson et al., 1999; de Lange et al., 2008), whereby the formation and enrichment of barite
crystals occurs in predominantly siliceous detritus during particle sinking through the water
column (Bishop, 1988; Dymond et al., 1992; Francois et al., 1995), and newly formed barite
crystals are imbedded in a matrix of organic particles. In line with our findings on δ15N and TOC
we consider it possible that the accumulation of biogenic barium in sapropels may also be
controlled by OM preservation. During particle disaggregation, intense OM remineralization and
cell lysis during particle sinking in the water column under normal oligotrophic conditions, barite
crystals become exposed to and dissolve in sea water (Dehairs et al., 1980; Dehairs et al., 1990).
This affects more than 70 % of the particulate Ba flux in the recent Southern Atlantic Ocean
(Dehairs et al., 1980). On the other hand, if OM is less degraded and remineralized during passage
56
δ15N in Mediterranean sediments
through a suboxic or anoxic water column where OM decay is inhibited, barite is shielded in more
or less intact detritus and accumulates in higher relative amounts in the sediments. This may have
been the case during Mediterranean sapropel formation and may explain the high Ba contents
without the need of enhanced organic carbon flux rates.
In the oxidized upper part of S1, however, Ba concentrations are decoupled from organic
carbon. Whereas OM is respired to CO2 and NH4+ during diagenesis, elevated barium
concentrations remain as a relatively inert trace of former OM rich layers (Thomson et al., 1995)
that outline the original extent of sapropels.
4.5. Conclusions
We have shown that variations of δ15N in Eastern Mediterranean sediment records are closely
tied to OM preservation and do not reflect the isotopic composition of the reactive nitrogen source
or nitrate utilisation. Diazotrophic nitrogen fixation very likely has been the dominant N source at
least during the Holocene and probably since the Pleistocene – regardless of whether the
Mediterranean Sea was in sapropel or non-sapropel mode. Spatial patterns in δ15N of S1 can also
be attributed to variable degrees of OM preservation due to differences in water column
oxygenation at different sites. TOC accumulation rates in long-term sediment trap moorings from
the EMS are the same as those calculated for sapropel S1, so that present day primary production
levels would be sufficient to create a S1 analog under conditions of deep water anoxia. The
strongest argument for elevated organic carbon export fluxes from the ocean surface during
sapropel times are elevated Ba/Al records, which also may be affected by water column
oxygenation.
57
58
5. Influence of Diagenesis on Sedimentary δ15N in the Arabian Sea Over the
Last 130 kyr
Abstract
We determined the ratio of stable nitrogen isotopes
15
N/14N (δ15N) and amino acid
composition in coeval sediments from Ocean Drilling Program (ODP) Hole 722 B in the
central Arabian Sea and from Hole 724 C situated on the Oman Margin in the western
Arabian Sea coastal upwelling area. The records span the last 130 kyr and include two glacialinterglacial cycles. These new data are used in conjunction with data available for surface
sediments and other cores from the northern and eastern Arabian Sea to explore spatial
variations in the isotopic signal and to evaluate effects of water depth, sedimentation rates and
oxygen content at the sea floor on the isotopic signal. All core records have similar patterns
through time, but different average values for δ15N at each core location. All records show a
positive correlation of δ15N and with an amino-acid derived Reactivity Index (RI; Jennerjahn
and Ittekkot, 1997) that indicates the relative degradation state of organic matter. The data
confirm that changes in the isotopic source signal of nitrate, commonly attributed to variations
in the intensity of denitrification in the regional mid-water oxygen minimum zone (OMZ),
covary with enhanced productivity, but also with enhanced organic matter preservation.
Different accumulation rates of the cores apparently have minor effect on amino acid
reactivity and δ15N. Hence, parallel but offset trends in δ15N and the RI at the different core
locations through time are interpreted to reflect the isotopic signature of source nitrate; these
original δ15N are altered during early diagenetic degradation of N-bearing organic matter and
can be normalized by using the RI. An evaluation of two preservation indexes based on amino
acid composition (RI and the Degradation Index, DI; Dauwe et al., 1999) in both recent
sediments and core samples suggests that the RI is more suitable than the DI in estimating the
state of organic matter degradation in older sediments.
59
Chapter 5
5.1. Introduction
The isotope ratio 15N/14N in sediment records, expressed as δ15N, is a commonly used proxy to
reconstruct changes in the oceanic nitrogen cycle and to evaluate past ocean productivity (Farrell
et al., 1995; Ganeshram et al., 2002; Altabet, 2007). The δ15N of sediments reflects nitrogen
sources (dissolved inorganic nitrogen, DIN) (Sweeney and Kaplan, 1980), as well as
transformation processes in the ocean that cause isotopic fractionation during cycling of nitrogen
(Altabet, 2006). Nitrate - the major DIN species – has an average δ15N value of about 5.0±0.5‰
in deep waters of the modern global ocean (Sigman et al., 2000), which may have been different
in the geological past in response to different source and sink terms (Brandes and Devol, 2002;
Deutsch et al., 2004). Nitrate in modern intermediate water masses has more variable isotopic
signatures than the deep reservoir which can be due to denitrification in oxygen-depleted
subsurface waters that enriches the residual nitrate in
15
N up to above 15‰ (e.g., Naqvi et al.,
1998; Voss et al., 2001; Thunell et al., 2004) or to the more recently discovered anammox process
that probably also enriches residual 15NO3- (Ward et al., 2009). Assimilation of such an enriched
source entering the euphotic zone leads to δ15N values of particulate organic matter and sediments
far above 5‰ (Naqvi et al., 1998; Altabet et al., 1999a; Voss et al., 2001). Isotopically depleted
δ15N values, on the other hand, often occur in sediments with a strong imprint of relatively
pristine terrestrial nitrogen supply (Gaye-Haake et al., 2005; Voss et al., 2005) or in areas of N2fixation, which regionally or seasonally introduces depleted δ15N that has the isotopic signal of air
(~0‰; Mariotti, 1984; Karl et al., 2002).
Isotopic fractionation during biological DIN uptake also has an effect on δ15N of sediments.
Lateral gradients of δ15N in particulate N assimilated from an isotopically homogeneous DIN
source are caused by “Rayleigh”-type fractionation during progressive DIN uptake by
phytoplankton. The initial products are isotopically depleted over the nitrate pool, whereas later
products are enriched over the original substrate (Altabet and Francois, 1994). This process can
explain increases of δ15N in sediments with increasing distance from nitrate sources such as river
mouths and upwelling areas (Holmes et al., 1997; Pichevin et al., 2005). Incomplete nitrate
utilization has been invoked to explain varying δ15N in sedimentary archives (Calvert et al., 1992;
Farrell et al., 1995).
Due to the short residence time of nitrogen in the ocean, the nitrogen cycle responds to
perturbations on very short time-scales (Tyrrell, 1999, Deutsch et al., 2004). Variations of δ15N
values in late Quaternary sediment records of the Arabian Sea have been interpreted as signals of
a response of the nitrogen cycle to climatic oscillations. Climatic changes led to changes in ocean
60
Influences on sedimentary δ15N in the Arabian Sea
circulation and induce changes in denitrification rates, assimilation rates, nutrient limitations, or
nitrogen fixation (Suthhof et al., 2001; Altabet et al., 2002; Ganeshram et al., 2002).
However, quantitative reconstructions of the global or regional N-cycles from geological
records are generally hampered by early diagenetic overprinting of the original
15
N/14N ratio.
Offsets in δ15N by 3-5‰ between particles in the water column and surface sediments have been
observed in open ocean sediments (Francois et al., 1992; Francois et al., 1997; Gaye-Haake et al.,
2005) and have been related to the intensity of degradation as reflected in the reactivity of amino
acid (AA) mixtures in sediments (Gaye-Haake et al., 2005). Although exact mechanisms remain
unknown, kinetic isotope fractionation during protein hydrolysis (Bada et al., 1989; Silfer et al.,
1992) and deamination (Macko and Estep, 1984), resulting in the preferential loss of 15N depleted
compounds during organic matter (OM) degradation, very likely causes this enrichment.
Decomposition of particulate OM in aquatic environments follows different pathways and
occurs in various steps. The main controlling factors are water column oxygenation, duration of
particle sinking as a consequence of water depth and particle size as well as the sealing efficiency
driven by sedimentation rates (e.g., Müller and Suess, 1979; Libes, 1992) that influence the
residence time of OM at the sediment water interface. In contrast, sedimentary degradation after
burial in the sediment is mediated by other microbial assemblages/communities than in the water
column and, depending on the depth of burial and sequestration time, by kinetic/thermal decay
ending in catagenesis (e.g., Tissot and Welte, 1978). Indexes of organic matter quality derived
from AA composition, such as the Reactivity Index RI (Jennerjahn and Ittekkot, 1997) and the
Degradation Index DI (Dauwe et al., 1999), are sensitive to degradation of OM in the water
column and at the sediment water interface and have often been used to estimate organic matter
quality (e.g., Unger et al., 2005; Kaiser and Benner, 2009); the extent to which they are
overprinted by late-stage diagenesis upon burial during several thousand years after sedimentation
has not been studied.
Our aim is to quantify diagenetic influences on sedimentary δ15N. We investigate surface
sediments that were deposited in a wide range of water depths and oxygen contents in bottom
waters, and sediment cores with variable sedimentation rates over the last 130 kyr from the
Arabian Sea. The main goal of our study is to quantify potential influences on sedimentary δ15N
other than source nitrate δ15N composition to permit an unbiased reconstruction of nitrate isotopic
composition in this key area of the global N-cycle (Altabet et al., 2002). For this we produced
paired data of δ15N and of indices of organic matter degradation based on AA composition (Cowie
and Hedges, 1992; Jennerjahn and Ittekkot, 1997; Dauwe and Middelburg, 1998; Dauwe et al.,
1999). To test the two indexes RI and DI for suitability in studies of OM degradation over
61
Chapter 5
geological periods we use statistical methods on data of amino acid composition in samples from
ODP cores 722 B and 724 C and derive an index that captures OM degradation in sediment cores
better than previously used indexes.
5.2. Study Sites
The data set of 46 surface sediments from the Arabian Sea north of 15° N and deeper than 250
m water depth published by Gaye-Haake et al. (2005) is our basis for estimating the effects of
water depth and bottom-water oxygen contents on both the degradation of amino acids and δ15N.
Oxygen contents in deep waters were interpolated for the sampling sites from the climatology of
Gouretski and Koltermann (2004).
Figure 5.1: Oxygen concentrations (ml/l) in bottom water of the Arabian Sea (Gouretsky and Koltermann, 2004)
and locations of sites referred to in the text. This map is produced by using Ocean Data View (Schlitzer, 2009).
62
Influences on sedimentary δ15N in the Arabian Sea
New data were raised from 2 Ocean Drilling Program (ODP) Sites (722 B and 724 C) in the
western Arabian Sea; additional core data are from cores 111 KL, NAST and EAST in the
northern and central Arabian Sea (Suthhof et al., 2001; Ivanova et al., 2003) (Figure 5.1 and Table
5.1). Site 722 is located on the Owen Ridge at a water depth of 2028 m (16°37’N; 059°48’E), i.e.
below the oxygen minimum zone of the Arabian Sea. Due to low detrital and biogenic particle
fluxes in this offshore setting, the bulk sediment accumulation rates (BAR) calculated from linear
sedimentation rates (based on the age model and on determinations of dry-bulk densities given in
Prell et al. (1989)) on the Owen Ridge are lower than at the Oman margin Site 724. Site 724 is
located on the upper slope of the Oman margin at a water depth of about 600 m (18°17’N;
057°28’E), i.e. within the oxygen minimum layer. Sediments at Site 724 have higher organic
carbon contents as well as higher accumulation rates compared to those at Site 722. Age models
for Holes 722 B and 724 C are based on oxygen isotope stratigraphies of Clemens and Prell
(1991) and Zahn and Pedersen (1991) and have been adjusted to visually fit the high-resolution
δ15N record in core NAST.
Core NAST was taken in the northern central Arabian Sea at 19°59.9’N and 065°41.0’E from
a water depth of 3170 m; core EAST was raised at 15°35.5’N and 068°34.9’E at 3820 m water
depth. Further details on these two cores are given in Suthhof et al. (2001), Ivanova et al. (2003)
and Guptha et al. (2005). δ15N of the NAST core older than 60 ky and the entire EAST core are
published here for the first time. The Pakistan margin core 111 KL was retrieved from 23°05.8’N
and 066°29.0’E at a water depth of 775 m and was described in Suthhof et al. (2001).
Table 5.1: Geographical position, water depth, and bottom water oxygenation (BWO) as well as examined depth
interval (downcore from seafloor in m) and time equivalent (kyr) of examined cores.
Core
ODP 722 B
ODP 724 C
111KL
EAST
NAST
Latitude
(°N)
Longitude
(°E)
Bottom depth
(m)
Core length
(m)
Time interval
(kyr)
BWO
(ml/l)
16°37'
18°17'
23°06'
15°36'
20°00'
059°48'
057°28'
066°29'
068°35'
065°41'
2028
603
775
3820
3170
5.23
12.26
14.00
5.00
5.60
7.1- 130.1
5.6- 143.9
5.3- 60.8
0.2- 132.9
0.1- 132.8
2.78
0.18
0.09
3.39
2.89
63
Chapter 5
5.3. Methods
5.3.1. Organic carbon and nitrogen
Total carbon and nitrogen were measured by a Carlo Erba elemental analyser 1500 (Milan,
Italy). The precision of this method is 0.2% for carbon and 0.02% for nitrogen. Carbonate
percentages were determined by a Wösthoff Carmhograph 6 (Bochum, Germany). The relative
error of carbonate analyses is 1%. Total organic carbon (TOC) was calculated as the difference
between total and carbonate carbon.
5.3.2. Stable nitrogen isotopic ratio (δ15N)
The ratio of the two stable isotopes of nitrogen (15N/14N) is expressed as δ15N after
determining the abundance of the two isotopes in samples (after combustion and reduction of NOx
to N2) by mass spectrometry:
δ15Nsample (‰) = (Rsample/Rstandard-1)*1000
R=15N/14N
(5.1)
15
The standard is atmospheric N2 (δ N = 0‰).
δ15N values were determined using a Finnigan MAT 252 gas isotope mass spectrometer after
high-temperature flash combustion in a Carlo Erba NA-2500 elemental analyzer at 1100°C. Pure
tank N2 calibrated against the reference standards IAEA-N-1 and IAEA-N-2 of the International
Atomic Energy Agency as well as a sediment standard that was used as a working standard.
Analytical precision was better than 0.1‰ based on replicate measurements of a reference
standard. Duplicate measurements of samples resulted in a mean standard deviation of 0.19‰
(Bahlmann et al., 2009).
5.3.3. Amino acids
Total hydrolysable AA were analysed with a Biochrom 30 Amino Acid Analyser after
hydrolysis of 30-40 mg of sediments with 6 N HCl for 22 hours at 110°C. After separation with a
cation exchange resin, the individual monomers were detected fluorometrically. Duplicate
analysis according to this method results in a relative error of 4% for total AA. A list of AAs
analyzed and the abbreviations used is given in Table 5.2. Further analytical details are given by
Jennerjahn and Ittekkot (1999).
64
Influences on sedimentary δ15N in the Arabian Sea
Table 5.2: Amino acids analyzed in this study, their abbreviations and occurrence in indexes. DI = Degradation
Index; RI = Reactivity Index; PCA = Principal Component Analysis.
Protein amino acids
Abbrev.
cysteic acid
taurine
aspartic acid
threonine
serine
glutamic acid
glycine
alanine
valine
methionine
iso-leucine
leucine
tyrosine
phenylalanine
histidine
tryptophane
lysine
arginine
Cya
Tau
Asp
Thr
Ser
Glu
Gly
Ala
Val
Met
Ile
Leu
Tyr
Phe
His
Trp
Lys
Arg
DI
x
x
x
x
x
x
x
x
x
x
x
x
x
RI PCA
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
Non-protein amino acids
ornithine
β-alanine
γ-aminobutyric acid
Orn
β-Ala
γ-Aba
x
x
x
x
x
Ratios and indices derived from the monomeric distribution of AA are used to estimate the state
of AA degradation:
Reactivity Index (RI)
The aromatic AAs tyrosine (Tyr) and phenylalanine (Phe) are concentrated in cell plasma and
are therefore most labile during cell lysis and decay processes (Hecky et al., 1973). Non-protein
β-alanine (β-Ala) and γ-aminobutyric acid (γ-Aba) occur only in trace amounts in living tissue but
become enriched as metabolic end-members during microbial decay (Cowie and Hedges, 1992).
Hence, the ratio of aromatic to non-protein AAs (the RI) reflects the relative degradation state of
OM (Jennerjahn and Ittekkot, 1997):
RI =
Tyr + Phe
β − Ala + γ − Aba
(5.2)
Poorest OM preservation is indicated by values close to 0; living marine plankton ranges between
4 and 6 (Jennerjahn and Ittekkot, 1997).
65
Chapter 5
Degradation Index (DI)
The DI (Dauwe et al., 1999) statistically evaluates relative abundances of the 14 most
common protein AAs (see Table 5.2). Dauwe et al. (1999) used the AA composition of 28
samples, representing a wide range of environmental settings and degradation state from living
plankton to Pleistocene sediments, as the data matrix for a principal component analysis (PCA).
The first PC factor scores derived from the analysis were considered to be the DI. To apply this
index to other datasets, the molar percentages of individual AAs have to be inserted as vari into
the following formula:
DI =
 var i − AVG var i 
 * fac.coefi
 STD var i
∑
i
(5.3)
where the deviation of the molar percentage of each AA from an average value (AVG) is
multiplied by an individual coefficient (fac.coef). AVG, fac.coef and STD (standard deviation) are
given by Dauwe et al. (1999). The sum of all operations yields the DI of a given sample; higher
values indicate better preservation than lower values. This approach has been successfully applied
to determine different stages of early OM decomposition in aquatic environments (e.g. Ingalls et
al., 2003; Gaye-Haake et al., 2005).
5.4. Results
5.4.1. Surface sediments
RI and DI show the same patterns in surface sediments but the DI was used for the same data
set by Gaye-Haake et al. (2005) to study the relationship between organic matter preservation and
the δ15N. Here we use the RI as it is more applicable for cores studies (see below) and correlate it
with δ15N, water depths, and oxygen in deep waters for the set of 46 surface sediments from the
earlier study of Gaye-Haake et al. (2005). The RI is significantly negatively correlated at the
99.9% significance level with both water depth (r2 = 0.82) (Figure 5.2a) and oxygen in bottom
waters (r2 = 0.61) (Figure 5.2c). Also, the correlation of δ15N with both water depth (r ² =0.67)
(Figure 5.2b) and oxygen (r² = 0.65) (Figure 5.2d) are statistically significant at the 99.9% level.
Similar correlations can be obtained by using the DI.
66
Influences on sedimentary δ15N in the Arabian Sea
Figure 5.2: Amino acid Reactivity Index (RI) and δ15N (‰) vs. water depth and bottom water oxygenation
(ml/l) in a set of 46 surface sediments. RI/ water depth (a), δ15N/ water depth (b), RI/ bottom water oxygenation
(c), δ15N/ bottom water oxygenation (d), δ15N/ RI (e), water depth/ bottom water oxygenation (f).
67
Chapter 5
5.4.2. Core data
The TOC contents at ODP Site 722 are between 0.3% in the older core section and 1.3% for
the youngest samples of an age of about 7 kyr. Total nitrogen ranges between 0.03 and 0.14%.
The accumulation rate at this site varied between 2.5 and 13.5 g*cm-2*kyr-1 over the last 130 kyr,
and the entire sequence analyzed has been deposited under oxygenated bottom water. The RI
ranges between 0.6 and 2.0 (Figure 5.3) and indicates poor to moderate preservation of amino
acids. The δ15N values are between 5.5 and 9.0‰ (Figure 5.3) and do not vary systematically with
either sediment accumulation rates or isotopic stage boundaries. Organic carbon contents in
sediments of Site 724 are between 0.1% and 2.7% and thus are higher than at Site 722 in the core
section younger than about 50 kyr and similar to Site 722 in the older core sections. TN
concentrations are between 0.03 and 0.3% with δ15N values between 3.8 and 8.0‰. Amino acid
concentrations are between 0.2 and 5.6 mg g-1 and track variations in TOC concentrations. The RI
varies between 1.3 and 3.6 (Figure 5.3), indicating moderate to good preservation. Two values in
RI are considered to be outliers (1.29 in Marine Isotope Stage (MIS) 1 and 3.34 in MIS 3) and are
excluded in later correlations. As in samples of Site 722, there is no systematic relationship of
δ15N with either sediment accumulation rates or isotopic stage boundaries. The average δ15N over
the last 130 kyr is 5.5‰ compared to 7.2‰ at Site 722. This offset of 1.7‰ is more or less
uniform through time between the two locations, with exceptional phases in the Holocene and
during MIS 3 when δ15N at Site 724 varies with large amplitude and the δ15N levels at the two
sites converge.
Variations in δ15N through time in NAST, EAST and 111 KL are very similar to those of ODP
722 B and 724 C series (Figure 5.4). δ15N ranges from 5.4‰ (NAST) and 5.7‰ (EAST) up to
10.6‰ and 10.2‰, respectively, so that the amplitudes of change are similar to that of 724 C, but
at a higher level of enrichment in 15N. Core 111 KL (Suthhof et al., 2001) only covers the last 60
kyr; here, at the Pakistan continental margin, δ15N varies between 4.2 and 9.8‰.
68
Figure 5.3: Bulk sediment accumulation rates (BAR), RI, and δ15N (‰) of samples from site 722 B (solid) and 724 C (dashed). MIS= Marine Isotope Stage; shaded are glacials.
Figure 5.4: δ15N (‰) of samples from Hole 724 C (dashed) and core EAST (solid) and Hole 722 B (dashed) and core NAST (solid) plus 111 KL (Suthhof et al., 2001).
Influences on sedimentary δ15N in the Arabian Sea
5.4.3. Comparison of amino acid indexes in the core studies
DI and RI diverge in samples from ODP 722 B (not shown) and 724 C from the Oman Margin
(Figure 5.5). While the RI traces the δ15N record, the relationship between δ15N and the DI is
ambiguous: The DI indicates the poorest preservation in the youngest samples, followed by a
continuous downcore increase towards a supposedly better preservation in the oldest samples. To
obtain independent biogeochemical indicators to the data set created here, we carried out a PCA
on amino acid data of core 722 B and 724 C using the mole-% values of all 21 individual AAs
analyzed (Table 5.2). The PCA yielded two factors explaining about 55% of the total variance.
The scores of the first factor plotted versus depth show an exponential downcore increase,
whereas the scores of the second factor oscillate in concert with the δ15N record (Figure 5.6). This
parallel trend in δ15N and the second factor score of AA data is evident in both data sets, but
whereas the correlation is highly significant (at the 99% level) for samples from 724 C, it is not in
722 B.
Figure 5.5: AA derived Degradation Index (DI) and Reactivity Index (RI) in ODP core 724 C show little
congruence in indication. For both indices higher values indicate better preservation than lower values.
71
Chapter 5
Figure 5.6: Variations in second PC factor scores (left plot) of AA PCA are similar to millennial scale
oscillations in δ15N (mid plot) and Ba/Al (data from Leuschner and Sirocko, 2003; right plot).
5.5. Discussion
5.5.1. Applicability of amino acid indexes in sediment cores
The fact that the RI and DI are comparable in surface sediments but diverge in core profiles
suggests that degradation processes differ depending on sediment age. Whereas microbial
degradation processes and subsequent physical segregation dominate in the water column and at
the sediment water interface, long-term abiotic processes (related to temperature, equilibrium- and
kinetic processes) in sediment cores alter the AA composition in a different way. Glycine (Gly)
and Threonine (Thr) are enriched during water column degradation as a result of their initially
high concentration in cell walls, preferentially accumulating in a more resistant OM fraction (Lee
and Cronin, 1984). For this reason, their enrichment is an indicator of OM degradation as
expressed in the DI. During degradation in the sediment column, when proteins are further
degraded and monomers are incorporated into geopolymers, these two AAs are depleted relative
to bulk AA content (Figure 5.7, Thr not shown). On the other hand, Phe and Glutamic acid (Glu)
(concentrated in cell plasma; Hecky et al., 1973) tend to be depleted during degradation in the
72
Influences on sedimentary δ15N in the Arabian Sea
water column and in surface sediments as a consequence of cell break-down (Lee and Cronin,
1984), but are relatively enriched downcore (Figure 5.7, Glu not shown).
Figure 5.7: Molar percentages of protein AA
glycine (Gly) and phenylalanine (Phe) in ODP
core 724 C. Relative abundance of both amino
acids due to depth and time is opposite as
expected from water column decay processes.
The relative abundance of Gly is schematically shown in Figure 5.8: after initial enrichment in
the water column, its relative contribution to total AA rapidly decreases in the sediments. This
may reflect preferential degradation by microbes utilizing AAs of low molecular weight, so that
larger AAs (such as Phe) may be relatively enriched over Gly in the course of degradation. The
later stage diagenetic effects are not captured by the DI. The index was derived from a broad suite
of surface sediments of different sedimentation and productivity regimes, but only few older
sediments from cores, mostly sapropels and turbidites (Dauwe et al., 1999) that possibly do not
represent undisturbed and continuous sedimentary degradation. This explains why the state of
degradation in sediments that have undergone downcore diagenesis may be not well reflected in
the DI; in our assessment, the DI is less applicable to characterize the state of degradation in the
sediment sequence from Hole 724 C.
73
Chapter 5
Figure 5.8: Schematic of Gly decay from the water column to the deeper sediment. In the water column and at
the sediment-water interface, Gly (as a cell wall constituent) becomes relatively enriched in a more stable OM
fraction. In the sediment Gly is rapidly consumed by benthic microorganisms due to its easy metabolization and
high availability. Further decay in the deeper sediment is achieved by slow microbial uptake and katagenesis.
Advanced degradation also affects the AAs used for the RI equation. In the water column the
non-protein AAs β-Ala and γ-Aba are relatively enriched over the aromatic AAs which has been
explained by the fact that non-protein AAs are either not utilizable for metabolism or are
decarboxylation products of aspartic acid and Glu, respectively, and are therefore accumulating
during diagenesis (Lee and Cronin, 1982). Non-protein AAs are thus enriched over the most labile
aromatic AAs Tyr and Phe (Jennerjahn and Ittekkot, 1997). In our cores, both non-protein and
total aromatic AAs show significant downcore enrichment relative to the other AAs. Tyr and Phe,
however, behave very differently. Whereas Phe gets twice as much enriched as β-Ala and γ-Aba,
74
Influences on sedimentary δ15N in the Arabian Sea
Tyr gets depleted close to the detection limit. These processes evidently lead to a conservation of
the RI originally achieved after degradation in the water column and at the sediment water
interface. As the RI of both ODP cores is correlated with the δ15N record (see Section 5.5.4.) and
with the Ba/Al record (as a proxy for paleoproductivity or preservation) from core KL 70
(Leuschner and Sirocko, 2003) located close to 722 C (Figure 5.6) we consider this indicator to be
more suitable for organic matter preservation in the water column and in surface sediments than
the DI, which appears to be overprinted by late stage OM degradation in sediment cores. In other
core studies, however, the DI appears to be robust (Chapter 4) highlighting the significance of
environment specific conditions on decay processes.
5.5.2. The relationship between organic matter preservation and δ15N: Influence of water
depth, bottom water oxygen concentrations and sediment accumulation rates
The paired data on AA preservation (RI) and δ15N of 46 surface sediments (Figures 5.2 a-f)
suggest a significant influence of water depth and oxygen concentrations on the degradation state
of organic matter and the N-isotope composition of sediments. This may reflect protein
degradation during passage of fresh organic matter through the water column and at the sediment
water interface where the degradation rates are high (Honjo et al., 1982; Cole et al., 1987). Recent
deep-sea sediments deposited in an oxic environment reveal the largest δ15N increase over sinking
particles (3-4‰; Francois et al., 1992; Francois et al., 1997; Altabet et al., 1999a), whereas δ15N
in sediments from suboxic continental margins only increase by <1‰ (Holmes et al., 2002;
Thunell et al., 2004; Gaye-Haake et al., 2005).
The change in organic matter composition with increasing water depth may reflect an
interplay of processes that are related to the distance to the coast. Near coastal environments are
characterized by higher productivity, shorter food chains and higher export production than
pelagic environments (Eppley and Peterson, 1979). The allochthonous supply of detritus from
land further enhances sedimentation and mass accumulation rates and thus promotes a relatively
fast sealing of organic matter in the sediment that reduces the period of intensive degradation at
the sediment-water boundary. Higher sinking speeds, induced by higher total fluxes and mineral
ballast, also reduce the time for degradation in the water column (Sirocko and Ittekkot, 1992).
More distal environments in contrast have lower primary production and bulk sedimentation rates
that are usually located in areas with a deeper water column. The decreasing productivity and
increasing degradation contribute to a gradient in organic matter flux away from land and with
increasing water depth (for a detailed discussion, including parts of this dataset see Suthhof et al.
(2000)). This general tendency is strengthened by the specific situation in the Arabian Sea.
Bottom water oxygen concentrations are lowest where the OMZ impinges on the shelf and slope.
75
Chapter 5
This additional effect of low oxygen levels further enhances organic matter preservation at
relatively shallow water depths and amplifies the differences in composition between shallow and
deep stations.
A comparison of average δ15N for the last 60 kyr between cores 722 B, 724 C, 111 KL, NAST
and EAST implies a slight increase with water depth converging to a δ15N value of about 8‰
below 3000 m depth (Figure 5.9).
Figure 5.9: Average δ15N (‰) of the study sites over the last 60 kyr versus water depth. Squares: recent surface
sediments. Both datasets show 15N enrichment due to increasing water depth.
In agreement with our findings on surface sediments, we interpret the increase in average δ15N
of cores with water depth as a result of (i) increased bottom water oxygenation from the slope to
the deep sea; (ii) effects related to the distance from land such as gradients in sedimentation rates
and primary production; and (iii) early diagenetic alterations related to the duration of particle
passage through the oxic water column. In contrast to Pichevin et al. (2007) our data do not reveal
a negative correlation of average sedimentation rates and the arithmetic mean of δ15N which may
be due to the much smaller range of sedimentation rates in our data set. Similarly, we do not find
a very significant relationship between sedimentary δ15N and RI with bulk accumulation rates
(BAR) over core records at ODP sites 722 B and 724 C (Figure 5.3). This may also be due to a
small range of BAR so that sealing effects on organic matter preservation may be masked by other
processes having a much stronger imprint on δ15N. Oxygen concentrations may probably have
76
Influences on sedimentary δ15N in the Arabian Sea
been more important for organic matter preservation in the sediment record, similar to the
findings in surface sediments (see above).
5.5.3. Other reasons for different δ15N values among the cores
A comparison of the δ15N record of 724 C with the deep-sea cores NAST and EAST shows a
constant offset of 2.4‰ in δ15N (as seen in surface sediments) over much of the Holocene and late
Pleistocene record (Figures 5.3 and 5.4). The δ15N values analyzed in core 722 B also correspond
very well to those at stations NAST and EAST, with similar values at δ15N minima and slightly
lower (1.5 to 2.5‰) for the maxima at these sites. Little offsets between the deep-sea cores may
result from different intensities in early diagenesis as discussed above. Generally, values of 722 B
are intermediate between the slope (111 KL, 724 C) and deep-sea cores (EAST, NAST). The
smaller amplitude in δ15N can be attributed to sedimentary winnowing effects that discriminate
smaller grain sizes (that usually are rich in OM) due to its elevated position on the Owen Ridge.
Recent studies indicate that the δ15NO3 of subsurface nitrate of the Arabian Sea is ~7-7.5‰
(Altabet et al., 1999b; Nagel et al., in prep.) and is generally homogenous in the entire basin.
While cores 722 B, EAST and NAST have significant fluctuations in δ15N in sediments older than
80 kyr and track the insolation record and millennial scale climatic fluctuations, the Oman Margin
record of Site 724 C displays uniformly low values. The generally lower δ15N in core 724 C
compared to present-day nitrate and the other cores, as well as the uniformly depleted values in
the lower part of core 724 C (Figure 5.4), can be explained by isotopic fractionation due to
incomplete nitrate assimilation since the location is situated in the Oman coastal upwelling zone
where nitrate is seldom exhausted.
5.5.4. Variations of denitrification and organic matter preservation over the last 130 kyr
Millenial scale oscillations of δ15N in the core records (Figures 5.3 and 5.4) confirm results of
earlier investigations on the same cores (Altabet et al., 1999b; Higginson and Altabet, 2004), and
are similar to other records from the region (Altabet et al., 1995; 2002; Suthhof et al., 2001).
Because of insufficient temporal resolution our data set aliases high-frequency variability seen in
other records - but clearly elevated δ15N characterize interglacial periods, whereas low δ15N
prevails during glacial periods and shorter cooling periods such as the Younger Dryas and
Heinrich events. This general pattern has been attributed to changes in denitrification intensity
which affects the isotopic composition of source nitrate (Altabet et al., 1995). Higher productivity
during enhanced monsoon-driven upwelling in conjunction with reduced ventilation of the
intermediate water masses is thought to have intensified the oxygen minimum zone, expanding
77
Chapter 5
the area and amount of denitrification, and thus raising the δ15N of sub-euphotic zone nitrate due
to a more intense denitrification (Altabet et al., 1995). Assimilation of this enriched nitrate pool
upon upwelling or after convection resulted in generally higher δ15N values of sediments
deposited during warm climatic stages. Temporal variation in δ15N in sediment cores thus mainly
reflects changes in the Arabian Sea nitrogen inventory (Altabet et al., 1995).
Figure 5.10: δ15N (‰) in comparison with amino acid reactivity (RI) in samples of Holes 722 B (n=51), Hole
724 C (n=58), 111 KL (n=52) (Suthhof et al., 2001), surface sediments (n=46) and sediment traps (n=6; mean
values). The shaded rectangle indicates the recent subsurface δ15NO3 of 7 to 7.5‰. Note that RI and δ15N are
positively correlated in core records but negatively correlated in surface sediments.
Exactly the same temporal patterns are evident in the second factor score of amino acid PCA
and in the RI of core 724 C as well as in the Ba/Al record of core 74 KL (Figure 5.6). Together,
the data imply that denitrification intensity (and δ15N of nitrate at the base of the mixed layer) is
closely coupled to higher primary production (providing substrate organic matter) and/or organic
matter preservation. In consequence, the relationship of δ15N and the RI in core samples is
opposite to the trend in surface samples - better preservation of organic matter is accompanied by
enriched δ15N (Figure 5.10). The same positive relationship in δ15N and RI also characterizes core
111 KL from the oxygen-deficient Pakistan margin environment (Suthhof et al., 2001) (Figure
5.10). Because the RI in surface sediments increases with decreasing bottom water oxygen
concentrations, the positive correlation between δ15N and RI found for each individual core seems
78
Influences on sedimentary δ15N in the Arabian Sea
to reflect changes in source nitrate composition rather than in degradation. We interpret the
positive correlation of RI and δ15N as a result of residual nitrate enrichment in
15
N due to the
intensity of denitrification. At the same time preservation becomes enhanced as a consequence of
contemporaneous oxygen limitation and higher primary production. Although the individual data
sets show some scatter, the regression lines of RI and δ15N for the three cores (Figure 5.10) have
approximately the same slopes determined by source nitrate isotopic composition, which in turn is
determined by denitrification intensity. The offsets between the three regression lines may be
caused by either differences in preservation or in the degree of nutrient utilization at the three
locations (see above). Highest δ15N and RI of the core records are in the range of values of recent
surface sediments and represent phases of intense denitrification.
The values for RI and δ15N in material from sediment traps (Figure 5.10) extend the trend of
surface sediments to fresh OM and are in the same range as the intersection of the slope of recent
PN (surface sediments and traps) with the δ15N of subsurface NO3-.
5.5.5. Estimating source nitrate isotopic composition
According to our considerations about processes that modulate spatial and temporal patterns
in the δ15N and RI of sediments in the Arabian Sea we suggest the following schematic summary
(Figure 5.11): (1) The δ15N of source nitrate is determined by the intensity of denitrification and
nutrient utilization in the euphotic zone, as well as by external inputs. This δ15N is transferred into
the particulate organic matter pool by assimilation. Spatial differences in particulate and settling
organic matter δ15N within the basin are produced when nutrients are incompletely utilized. (2)
There is a systematic relationship between the intensity of water column denitrification that
determines the δ15N of source nitrate and the preservation of organic matter in the core record of
the recent geological history. This is reflected by the typically positive correlation of δ15N and the
AA-derived biogeochemical indicator RI. (3) Diagenetic enrichment of the original δ15N of the
assimilated nitrate increases with water depth, bottom water oxygenation and other basin effects,
as indicated by a negative correlation of the RI and δ15N in surface sediments.
79
Chapter 5
Figure 5.11: Schematic interpretation of the relationship betweens sedimentary δ15N and RI in the Arabian Sea
as shown in figure 10: External sources, the intensity of denitrification and the degree of NO3- assimilation
determine the general setting of δ15N (1). Water column denitrification leads to an enrichment in 15N and
contemporaneous better preservation (2). During progressive OM decay (expressed by decrease in RI) δ15N
becomes enriched (3). The offset between the dashed arrow “water column denitrification (2)” and the solid
“sediment record“ represents the early diagenetic enrichment of the examined sediment cores.
In earlier studies the proportion of diagenetic enrichment on sedimentary δ15N has been
quantified by analyzing the δ15N preserved in diatom frustules (Sigman et al., 1999), foraminifer
shells (Ren et al., 2009) and biomarkers resistant to diagenesis (Sachs and Repeta, 1999). Here we
suggest a reconstruction of the pristine diagenetically non altered δ15N under the assumption of a
linear relationship between 15N enrichment and OM degradation as reflected in AA composition.
The diagenetic increase (labelled 3 in Figure 5.11) depicts the relative trends caused by different
processes that affect δ15N and RI as seen in surface sediments. This diagenetic offset should be
applicable to any other time slice in core records from the Arabian Sea. Assuming that the slope
of the regression has not changed significantly, we may use the δ15N-RI correlation obtained for
surface sediments (Figure 5.10) to reconstruct the δ15N of primary produced OM: Each δ15N value
(δ15Nsample) is corrected by the slope of δ15N vs. RI in surface sediments (δ15N= -0.67 RI =
msurf.sed.; Figure 5.10) to an assumed RI value before diagenetic enrichment of δ15N (RIoriginal) by
the equation:
δ15Noriginal = δ15Nsample + (msurf.sed.*∆RIoriginal-sample)
(5.4)
According to the intersection of surface sediments and sediment traps slope with the δ15N of 7
to 7.5‰ in recent subsurface NO3 we assign the RI of newly produced, isotopically non altered
80
Influences on sedimentary δ15N in the Arabian Sea
OM to ~ 5 (Figure 5.10, arrow). This value is in good agreement with data for living marine
plankton from Jennerjahn and Ittekkot (1997).
The main uncertainty of this reconstruction is the assumption that the slope of the RI-δ15N
regression has been stable over the geological history. A study carried out on recent sediments
from the South China Sea suggests that the slope is less steep under more oligotrophic conditions
(Gaye et al., 2009).
Figure 5.12: Reconstructed δ15N record for
primary produced OM in cores 722 B (gray
dashed) and 724 C (black). Compare to
Figure 3, right plot. The reconstructed
records show significant lower values and
records converge by 0.5‰ on average.
Despite these uncertainties we reconstructed the presumed pre-diagenetic δ15N of cores 724 C
and 722 B that generally have lower δ15N values, but larger amplitudes than the uncorrected time
series (Figure 5.12). The means of reconstructed δ15N-records converge by 0.5‰ and values are
almost the same during phases of intense denitrification. The calculated δ15N for isotopically
enriched interstadial sediments remain in a similar order of magnitude as recent NO3- (maximum
6.9‰ in both cores). However, for sediments deposited during glacial stages and Younger Dryas/
Heinrich events, reconstructed δ15N values show a significant offset between both cores (minima
81
Chapter 5
in δ15N= 1.3‰ for 724 C and 2.7‰ for 722 B). Generally, these very low values suggest intense
N2 fixation by cyanobacteria. This is feasible as enhanced inputs from Tibet supplied Fe/Mo (Karl
et al., 2002) as a consequence of a strengthened NE-monsoon during glacial stages (Tamburini et
al., 2003).
5.6. Conclusions
The fluctuations in δ15N in the sediment cores investigated in this study are not primarily
driven by diagenesis but by differences in δ15N of the assimilated nitrogen. The intensified OMZ
during the warm phases such as the Holocene and Interstadials is reflected by enriched δ15N
values, and this organic matter is less degraded than the material deposited during the colder
stadials, especially during the Younger Dryas and Heinrich events. Near shore cores (724 C, 111
KL) show similar pattern as deep Arabian Sea cores (722 B, NAST, EAST). However, the
absolute δ15N records are enriched by about 0.6-3.0‰ during both cold and warm stages in deep
sea cores compared to shallow cores. This offset is caused by (i) basin effects as there are
gradients in sedimentation rate, primary production, water depth and bottom water oxygenation
that lead to gradients in early diagenetic alterations in δ15N and (ii) higher fractionation during
NO3- uptake in the upwelling environment of core 724 C.
Established AA indices RI and DI diverge and are not comparable in the examined cores. We
suspect that the reason for this discrepancy lies in different pathways or rates of AA
decomposition in older sediments compared to decay in the water column and at the sediment
water interface; this difference is inherent in the calculation of AA quality by the RI, but is not
captured in the DI. A normalization of δ15N in core records to diagenetic non altered values of
primary produced OM by using the RI seems possible but vague. This reconstruction suggests that
the interstadial conditions were comparable to the recent late Holocene whereas cold stages were
characterized by significant nitrogen fixation stimulated by enhanced dust supply from the
Tibetian Plateau.
82
6. Conclusions and Outlook
6.1. Conclusions
This thesis successfully uses δ15N as a tracer of reactive N sources in the present-day
Mediterranean Sea and investigates factors determining the δ15N of particulate OM as well as the
applicability of δ15N as a proxy for N sources in sediment cores from the Mediterranean Sea and
the Arabian Sea.
Unusually depleted 15N values in all compartments of reactive N in the EMS are indicative for
the isolation of the basin from the global ocean. The low level of δ15N reflects the dominance of
isotopically depleted N sources and the lack of processes that enrich nitrate in
15
N. Isotopic
enrichment of TRN and suspended PN over NO3- and model data as well as the nitrate isotope
anomaly ∆(15,18) suggest that N-remineralisation and atmospheric deposited NOx are important
contributors of
15
N depleted reactive N to the mixed layer. Atmospheric input (mainly
anthropogenic) of 15N depleted NOx possibly decreases the δ15NO3- in deepwater during the last
decades.
Large scale variations in sedimentary δ15N-records of the Eastern Mediterranean during the
Holocene (including sapropel S1) are mainly driven by preservation, and not by the isotopic
composition of the reactive N source, or the degree of nitrate utilization. The enhanced
accumulation of OM in sapropel S1 can be explained by better preservation as a consequence of
deep water anoxia, without the need of elevated primary production.
Both studies will be important contributions to the long standing debates about the origin of
Mediterranean sapropels as well as the applicability of δ15N as a proxy, and in the evaluation of
current anthropogenic impacts on marine ecosystems.
Results of the Arabian Sea study, however, are less conclusive. Although my data confirm a
strong early diagenetic imprint on δ15N, and imply that the RI can be used to reconstruct the
primary produced δ15N, conclusions about the N inventory and utilization during the investigated
period are less clear. Opposed to the Mediterranean Sea study, variations in δ15N in the sediment
records from the Arabian Sea appear to be primarily driven by differences in δ15N of the
assimilated nitrate. Warm phases, such as the Holocene and Interstadials, are reflected by enriched
δ15N values and well preserved OM as a consequence of an intensified oxygen minimum zone in
intermediate water masses. Core records show similar basin-wide millennial-scale oscillations in
δ15N with an offset from one location to the next that I attribute to different degrees of early
diagenetic enrichment in δ15N and differences in NO3- availability at each core location.
83
Chapter 6
EMS and Arabian Sea cores are both archives of late Pleistocene to Holocene climate
variability in the monsoon system. They both show millennial-scale oscillations in TOC
accumulation and δ15N that are driven by monsoon intensity and follow Milankovitch cycles. The
signal in δ15N and its relationship to OM preservation, however, has an opposite sign in these two
settings, because better OM preservation goes in concert with depleted δ15N values in the EMS,
but with enriched δ15N in the Arabian Sea. This is because different environmental settings force
different processes that in the end determine the records. This opposite behavior illustrates that
single data of δ15N can be easily misinterpreted – but in conjunction with other proxies, δ15N is a
highly potent proxy.
6.2. Outlook
While studies performed towards this Ph.D.-thesis gave many fascinating results and some
answers, they also pose new questions and suggest further research and different approaches:
The monitoring of deep water NO3- isotopic composition in the Mediterranean Sea should be
continued to further investigate the current trend of
15
N depletion. Based on a broader data set
local models of Northern Ionian Sea stations and pelagic stations could be extended to a regional
Eastern Mediterranean isotope budget. An improvement of analytical techniques to determine
isotopic composition of small amounts of nitrate may allow to investigate the NO3- starved surface
layer in more detail, thus closing the gap between atmospheric deposition and the deep-water
nitrate pool to better trace nitrification in the nutrient-starved surface layer and the input of
anthropogenic reactive N. In my opinion, the EMS is an ocean basin (albeit a small one) where
(an as yet undefined) natural N-cycle is currently being flooded by anthropogenic reactive
nitrogen. But to fully understand only the present-day N-cycle in the EMS, we need additional
insights from inter-annual time-series of N-export flux, which currently are discontinuous, and
preferably from depths immediately below the thermocline at several stations. I also speculate that
TRN, which was
15
N depleted in our samples, should have a distinct seasonal cycle with
characteristic enrichment in summer months. The largest issue to date is the presence or absence
of significant N2 fixation, whether near the surface or in the enigmatic deep chlorophyll
maximum, and its true magnitude. And a nagging questions remains: If atmospheric inputs are
indeed large and have been delivering excess nitrate in the last decades, and if these inputs have
high (~50) N:P ratios as suggested by measurements, why have N:P ratios in deep water not been
rising?
84
Conclusions and Outlook
Further investigations on OM decomposition and N-isotope fractionation will be conclusive in
order to substantiate the findings of Chapter 4: Planktonic foraminifer shells in Eastern
Mediterranean sediments probably retain the original, primary produced δ15N signal in their
proteinaceous linings (Ren et al., 2009) and may (or may not) verify the DI-normalized values
presented in this study. An undisturbed sediment core from a permanently anoxic brine basin of
the EMS should reflect relative changes in the δ15N of primary produced OM that are not
overprinted by the diagenetic enrichment occurring today at the sediment-water-interface (e.g.
core #569); if there were changes in sedimentary δ15N deposited during the last decades it should
be due to anthropogenic perturbations but not to a diagenetic overprint. Studies on sinking
material and sediment cores from the Black Sea (as an analogue to Mediterranean sapropel
conditions) can in addition clarify the relationship of OM accumulation, N-isotope distribution
and Ba/Al ratios to preservation.
The evident linear to logarithmic relationship of δ15N and N (%) in core #569 and other
Mediterranean cores suggests an isotopic enrichment of the residual fraction during N-removal in
the sediments following a Rayleigh type fractionation. Assuming that variations of δ15N in
Mediterranean sediments of the Holocene were solely driven by differences in degradation from
the same amount of N with a similar δ15N by ammonification, it should be possible to calculate
the fractionation factor 15ε of ammonification in sediments. Results from coupled analyses of δ15N
in pore water ammonium and nitrate as well as organic N can be used to verify results of such
calculations.
The Arabian Sea study revealed that knowledge about amino acid decay under differing
environmental conditions and during progressive diagenesis is scarce and needs further research.
If the approach of correcting sedimentary δ15N for diagenetic increase by using amino acid spectra
to determine preservation is validated by other studies from areas with variable OMZs, modelled
estimates for the extent of water column denitrification and hence the rates for removal of reactive
N from the global ocean and emissions of N2O can be improved for older geological time. Again,
a comparison to δ15N records from planktonic foraminifer shells from sampling locations in
different water depths would be conclusive. Another interesting question arises from the current
debate whether denitrification or anammox is the main actor of N removal in the Arabian Sea
(Ward et al., 2009): Is the systematic relationship between OM preservation and 15N enrichment
the same for both, denitrification and anammox? And if it is not, can we distinguish the
proportion of both in core records via different slopes in the relationships between δ15N and
amino acid preservation indexes?
85
Acknowledgements
I would like to thank Prof. Dr. Kay-Christian Emeis for the supervision of my thesis; for his open
mind regarding new ideas and for valuable discussions. I further thank for organizing four
years of financial support and for the opportunity to participate in numerous conferences where
I could smell the breath of international science.
I thank Dr. Birgit Gaye for co-supervision, and together with Dr. Andrea Wieland and Prof. Dr.
Gerhard Schmiedl for endless scientific discussions that resulted in new ideas and
encouragement.
Many thanks to laboratory security agent Dr. Niko Lahajnar for never losing his patience about
my chaotic lab management, for scientific discussions and for support whenever I needed.
A special thank goes to Marc Metzke and Frauke Langenberg for their great support in sample
preparation, management and analyses.
Thanks to Andreas Neumann for scientific discussions and the nice office atmosphere.
Thanks to Birgit Nagel for supervision and support in the GKSS labs.
I would like to acknowledge the nice working and social atmosphere: Thanks to all my colleagues
of the IfBM, GKSS, the MERF project and during three research cruises!
Finally I thank my parents, brothers and sisters, my flat sharing in St. Georg and all my friends
who made my life easier during the past four years by giving me support or diversion.
86
Figure Captions
Figure 1.1: The marine nitrogen cycle
Figure 1.2: Rayleigh fractionation during NO3- assimilation in a closed system
Figure 3.1: Map of stations occupied in the Eastern Mediterranean Sea during r/v Meteor cruise
71-3 (January-February, 2007). NIS stations referred to in the text are stations H07 to
H12. The black square SE of Crete (water columns sampling station Ier1) marks the
location of the sediment trap deployment station (MID).
Figure 3.2: Profiles of fluorescence (a; arbitrary units), nitrate (b) and phosphate (c)
concentrations in the upper 400 m at 2 stations representative of NIS (H07) and
pelagic stations (Her03) show stratification between 80 and 230 m water depth and
indicate the biologically active mixed layer. An ongoing phytoplankton bloom in the
northern Ionian Sea (at station H07) is sustained by nitrate and phosphate provided
from ongoing regional thermocline deepening, whereas station Her03 illustrates the
mature and thick mixed layer with very low nutrient concentrations at pelagic sites.
Figure 3.3: Concentrations of nitrate (a), phosphate (b), TRN (c) and particulate N in total
suspended solids (d) plotted against water depth for all stations.
Figure 3.4: Isotopic composition δ15NO3- (a), δN18O3- (b), δ15TRN (c), and δ15PN (d) plotted
against water depth for all stations.
Figure 3.5: Composite seasonal diagram of SPN fluxes (squares) and δ15SPN (circles) and their
standard deviations at 2600 m water depth at station Ierapetra (34°26´N, 26°11´E,
water depth 3750 m). Fluxes (squares, black line and s.d. in in red) and δ15SPN
(circles, grey line, s.d. in blue) for three deployment periods (MID-1: 01/30/1999 to
04/13/1999), MID-2: 11/05/2001 to 04/01/2002 and MID-3: 01/30/2007 to
09/05/2007) at 2700 m water depth have been assembled in a surrogate annual cycle.
Figure 3.6: Plots of log-normal nitrate concentrations (X-axis) versus δ15NO3- (a) and δN18O3- (b)
found in samples in and above the nitraclines for individual station. The solid
regression line in Figure 6a is for samples in the northern Ionian Sea (NIS) where
87
Figure captions
nitrate is replenished from thermocline deepening (r2=0.89). In these samples,
15
N in
nitrate is enriched in the course of ongoing nitrate assimilation with an apparent
separation factor 15ε close to -3‰. In samples from the open Ionian Sea and Herodotus
Basin (pelagic stations), where the mixed layer was thick and nitrate poor at the time
of sampling, ln(nitrate) and δ15NO3- are uncorrelated. In Figure 3.6b, the correlation
between ln(nitrate) and δN18O3- for NIS samples is also significant (r2=0.78), with an
apparent separation factor
18
ε close to -5‰. There is no significant correlation in
samples from pelagic stations and from the northern Aegean station.
Figure 3.7: Depth plot of ∆(15,18) of nitrate for different station sets in the upper 500 m of the
water column.
Figure 4.1: Sampling sites in the Eastern Mediterranean Sea. Large dots: Sediment cores;
Triangle: Ierapetra Deep sediment traps and multicores; small dots: surface sediment
samples. The color shading is δ15N (‰) in surface sediments based on this study and
Struck et al. (2001). The map is produced by using Ocean Data View (Schlitzer,
2009).
Figure 4.2: δ15N vs. degradation index (DI) in cores MC 284/285 from Ierapetra Deep and mean
values from MID sediment traps. Insert: High resolution records of δ15N and DI of
multicores MC 284 and MC 285.
Figure 4.3: Downcore plots of total organic carbon (%TOC), barium to aluminum ratios,
degradation index (DI) and δ15N in multicores #569 (a), #562 (b) and #563 (c). S1
sapropel intervals and the burn-down zone above S1 are marked.
Figure 4.4: δ15N vs. DI in core #569 and mean values from MID sediment traps. Insert: δ15N vs. N
(%) in core #569. Residual N becomes isotopically enriched during remineralization.
Surface sediments and protosapropel are marked in both plots.
Figure 4.5: δ15N vs. DI for S1, S5 and S6 sapropels, core #569 and all Eastern Mediterranean
surface samples. Mean values.
Figure 5.1: Oxygen concentrations (ml/l) in bottom water of the Arabian Sea (Gouretsky and
Koltermann, 2004) and locations of sites referred to in the text. This map is produced
by using Ocean Data View (Schlitzer, 2009).
88
Figure captions
Figure 5.2: Amino acid Reactivity Index (RI) and δ15N (‰) vs. water depth and bottom water
oxygenation (ml/l) in a set of 46 surface sediments. RI/ water depth (a), δ15N/ water
depth (b), RI/ bottom water oxygenation (c), δ15N/ bottom water oxygenation (d),
δ15N/ RI (e), water depth/ bottom water oxygenation (f).
Figure 5.3: Bulk sediment accumulation rates (BAR), RI, and δ15N (‰) of samples from site 722
B (solid) and 724 C (dashed). MIS= Marine Isotope Stage; shaded are glacials.
Figure 5.4: δ15N (‰) of samples from Hole 724 C (dashed) and core EAST (solid) and Hole 722
B (dashed) and core NAST (solid) plus 111 KL (Suthhof et al., 2001).
Figure 5.5: AA derived Degradation Index (DI) and Reactivity Index (RI) in ODP core 724 C
show little congruence in indication. For both indices higher values indicate better
preservation than lower values.
Figure 5.6: Variations in second PC factor scores (left plot) of AA PCA are similar to millennial
scale oscillations in δ15N (mid plot) and Ba/Al (data from Leuschner and Sirocko,
2003; right plot).
Figure 5.7: Molar percentages of protein AA glycine (Gly) and phenylalanine (Phe) in ODP core
724 C. Relative abundance of both amino acids due to depth and time is opposite as
expected from water column decay processes.
Figure 5.8: Schematic of Gly decay from the water column to the deeper sediment. In the water
column and at the sediment-water interface, Gly (as a cell wall constituent) becomes
relatively enriched in a more stable OM fraction. In the sediment Gly is rapidly
consumed by benthic microorganisms due to its easy metabolization and high
availability. Further decay in the deeper sediment is achieved by slow microbial
uptake and katagenesis.
Figure 5.9: Average δ15N (‰) of the study sites over the last 60 kyr versus water depth. Squares:
recent surface sediments. Both datasets show 15N enrichment due to increasing water
depth.
89
Figure captions
Figure 5.10: δ15N (‰) in comparison with amino acid reactivity (RI) in samples of Holes 722 B
(n=51), Hole 724 C (n=58), 111 KL (n=52) (Suthhof et al., 2001), surface sediments
(n=46) and sediment traps (n=6; mean values). The shaded rectangle indicates the
recent subsurface δ15NO3 of 7 to 7.5‰. Note that RI and δ15N are positively
correlated in core records but negatively correlated in surface sediments.
Figure 5.11: Schematic interpretation of the relationship betweens sedimentary δ15N and RI in the
Arabian Sea as shown in figure 10: External sources, the intensity of denitrification
and the degree of NO3- assimilation determine the general setting of δ15N (1). Water
column denitrification leads to an enrichment in
15
N and contemporaneous better
preservation (2). During progressive OM decay (expressed by decrease in RI) δ15N
becomes enriched (3). The offset between the dashed arrow “water column
denitrification (2)” and the solid “sediment record“ represents the early diagenetic
enrichment of the examined sediment cores.
Figure 5.12: Reconstructed δ15N record for primary produced OM in cores 722 B (gray dashed)
and 724 C (black). Compare to Figure 3, right plot. The reconstructed records show
significant lower values and records converge by 0.5‰ on average.
90
Table Captions
Table 1.1: Fractionation factors ε for major transformation processes in the marine N-cycle. 14N
preference is expressed by negative values.
Table 1.2: Isotopic signatures of selected N-pools.
Table 3.1: Comparison of averages in water above, in and below the nitracline for samples taken
in stations of the Northern Ionian Sea (6 stations, Table 3.1a), at pelagic stations (10
stations, Table 3.1b) and at one station in the northern Aegean Sea (Table 3.1c) in the
EMS in January/February 2007 during Meteor expedition 71-3.
Table 3.2: Composite seasonal course of N-flux and δ15N of sinking particulate matter at station
MID (Ierapetra), water depth 2700 m. Data are from three separate and discontinuous
deployment periods starting in 1999 and ending in 2007, as explained in the text.
Table 3.3: Estimate of reactive N-inventories of the EMS in different depth intervals and massweighted δ15N of the different components (nitrate, particulate nitrogen PN, and total
reduced nitrogen, TRN). We calculated inventories (given in gigamol N) based on
interval water volumes for the EMS without Adriatic and Aegean sub-basins (R.
Grimm, pers. comm., 2009) and weight δ15N values by average concentrations found
during our cruise. The last column is the integrated and mass-weighted δ15N over all
components of reactive N for each interval, the last line are the integrated inventories
and δ15N of the entire water column.
Table 3.4: Calculation of the isotopic composition of residual nitrate (δ15N and δ18O) and
instantaneous and accumulated products of nitrate assimilation expected from
Rayleigh-fractionation in closed systems (NIS and pelagic mixed layers) with starting
concentrations and isotopic compositions corresponding to thermocline nitrate from
Table 3.1. Note that the permil fractionation factors 15ε and 18ε have been set to -5‰.
Results of measurements are also shown.
Table 4.1: Sediment trap moorings in the Ierapetra Deep: Position, water depth, sampling depth,
deployment period and trap type.
91
Table captions
Table 4.2: Sediment sampling locations and sample types: MC = Multicore, GC = Gravity Core,
ODP = Ocean Drilling Program Core.
Table 4.3: Sediment trap moorings in the Ierapetra Deep: Mean values of δ15N, DI, TOC and
total flux.
Table 4.4: δ15N, DI and TOC values of recent surface sediments as well as from S1, S5 and S6
sapropel interval (1)=estimated values; 2)=mean values of entire sapropel).
Table 5.1: Geographical position, water depth, and bottom water oxygenation (BWO) as well as
examined depth interval (downcore from seafloor in m) and time equivalent (kyr) of
examined cores.
Table 5.2: Amino acids analyzed in this study, their abbreviations and occurrence in indexes.
DI = Degradation Index; RI = Reactivity Index; PCA = Principal Component
Analysis.
92
List of abbreviations
AA = Amino Acid
DI = Degradation Index
DON = Dissolved Organic Nitrogen
ε = isotope fractionation factor
EMDW = Eastern Mediterranean Deep Water
EMS = Eastern Mediterranean Sea
GC = Gravity Core
LIW = Levantine Intermediate Water
MAW = Modified Atlantic Water
MC = MultiCore
MID = Mediterranean Ierapetra Deep
NIS = Northern Ionian Sea
OM = Organic Matter
OMZ = Oxygen Minimum Zone
PN = Particulate Nitrogen
R = isotope Ratio
RI = Reactivity Index
SL = gravity core
SPN = Sinking Particulate Nitrogen
SPON = Suspended Organic Nitrogen
TDN = Total Dissolved Nitrogen
TN = Total Nitrogen
TRN = Total dissolved Reduced Nitrogen
TOC = Total Organic Carbon
93
References
Altabet, M.A., 1988. Variations in nitrogen isotopic composition between sinking and suspended
particles: implications for nitrogen cycling and particle transformation in the open ocean.
Deep Sea Research I, 535-554.
Altabet, M.A. and Francois, R., 1994. Sedimentary Nitrogen Isotopic Ratio as a Recorder for
Surface Ocean Nitrate Utilization. Global Biogeochemical Cycles 8, 103-116.
Altabet, M.A., Francois, R., Murray, D.W. and Prell, W.L., 1995. Climate-related variations in
denitrification in the Arabian Sea from sediment 15N/14N ratios. Nature 373, 506-509.
Altabet, M.A., 1996. Nitrogen and carbon isotopic tracers of the source and transformation of
particles in the deep-sea. In: V. Ittekkot, P. Schafer, S. Honjo and P.J. Depetris (Editors),
Particle Flux in the Ocean. Wiley, London, pp. 155-184.
Altabet, M.A., Pilskaln, C., Thunell, R., Pride, C., Sigman, D., Chavez, F. and Francois, R.,
1999a. The nitrogen isotope biogeochemistry of sinking particles from the margin of the
Eastern North Pacific. Deep Sea Research I 46, 655-679.
Altabet, M.A., Murray, D.W. and Prell, W.L., 1999b. Climatically linked oscillations in Arabian
Sea denitrification over the past 1 m.y.: Implications for the marine N cycle
Paleoceanography 14, 732-743.
Altabet, M.A., Higginson, M.J. and Murray, D.W., 2002. The effect of millennial-scale changes in
Arabian Sea denitrification on atmospheric CO2. Nature 415, 159-162.
Altabet, M.A., 2006. Isotopic tracers of the marine nitrogen cycle: Present and past. Marine
Organic Mattter: Biomarkers, Isotopes and DNA 2, 251-293.
Altabet, M.A., 2007. Constraints on oceanic N balance/imbalance from sedimentary N-15 records.
Biogeosciences 4, 75-86.
Anastasakis, G.C. and Stanley, D.J., 1984. Sapropels and organic-rich variants in the
Mediterranean: sequence development and classification. In: D.A.W. Stow and D.J.W.
Piper (Editors), Fine-grained Sediments: Deep-water Processes and Facies. Geological
Society, London, pp. 497-510.
Anastasakis, G.C. and Stanley, D.J., 1986. Uppermost Sapropel, Eastern Mediterranean Paleoceanography and Stagnation. National Geographic Research 2, 179-197.
Antoine, D., Morel, A. and André, J.-M., 1995. Algal pigment distribution and primary
production in the eastern Mediterranean as derived from coastal zone color scanner
observations. Journal of Geophysical Research 100(C8), 16193-16209.
Arnaboldi, M. and Meyers, P.A., 2006. Patterns of organic carbon and nitrogen isotopic
compositions of latest Pliocene sapropels from six locations across the Mediterranean Sea.
Palaeogeography, Palaeoclimatology, Palaeoecology 235, 149-167.
94
References
Bada, J.L., Schoeninger, M.J. and Schimmelmann, A., 1989. Isotopic Fractionation During
Peptide-Bond Hydrolysis. Geochimica et Cosmochimica Acta 53, 3337-3341.
Badr, O. and Probert, S.D., 1993. Environmental Impacts of Atmospheric Nitrous-Oxide. Applied
Energy 44(3), 197-231.
Bahlmann, E., Bernasconi, S.M., Bouillon, S., Houtekamer, M., Korntheuer, M., Langenberg, F.,
Mayr, C., Metzke, M., Middelburg, J.J., Nagel, B., Struck, U., Voß, M. and Emeis, K.-C.,
Performance evaluation of nitrogen isotope ratio determination in marine and lacustrine
sediments: An inter-laboratory comparison. Organic Geochemistry, in press, corrected
proof.
Barber, R.T., Marra, J., Bidigare, R.C., Codispoti, L.A., Halpern, D., Johnson, Z., Latasa, M.,
Goericke, R. and Smith, S.L., 2001. Primary productivity and its regulation in the Arabian
Sea during 1995. Deep Sea Research II 48, 1127-1172.
Benitez-Nelson, C.R., 2000. The biogeochemical cycling of phosphorus in marine systems. EarthScience Reviews 51(1-4), 109-135.
Berman-Frank, I., Yogev, T., Aharonovich, D. and Beja, O., 2007. Nitrogen fixation in the
Eastern Mediterranean Sea, Oral presentation, IUGG XXIV General Assembly, Perugia,
Italy, PS001.
Berman, T. and Bronk, D.A., 2003. Dissolved organic nitrogen: a dynamic participant in aquatic
ecosystems. Aquatic Microbial Ecology 31, 279-305.
Béthoux, J.P. and Copin-Montegut, G., 1986. Biological fixation of atmospheric nitrogen in the
Mediterranean Sea. Limnology and Oceanography 31, 1353-1358.
Béthoux, J.P., 1989. Oxygen-Consumption, New Production, Vertical Advection and
Environmental Evolution in the Mediterranean Sea. Deep Sea Research I 36, 769-781.
Bishop, J.K.B., 1988. The barite-opal-organic carbon association in oceanic particulate matter.
Nature 332, 341-343.
Böhlke, J.K., Mroczkowski, S.J. and Coplen, T.B., 2003. Oxygen isotopes in nitrate: new
reference materials for 18O:17O:16O measurements and observations on nitrate-water
equilibration. Rapid Communications in Mass Spectrometry 17, 1835-1846.
Bourbonnais, A., Lehmann, M.F., Waniek, J.J. and Schulz-Bull, D.E., 2009. Nitrate isotope
anomalies reflect N2 fixation in the Azores Front region (subtropical NE Atlantic). Journal
of Geophysical Research 114, doi:10.1029/2007JC004617
Brandes, J.A., Devol, A.H., Yoshinari, T., Jayakumar, D.A. and Naqvi, S.W.A., 1998. Isotopic
composition of nitrate in the central Arabian Sea and eastern tropical North Pacific: A
tracer for mixing and nitrogen cycles. Limnology and Oceanography 43(7), 1680-1689.
Brandes, J.A. and Devol, A.H., 2002. A global marine-fixed nitrogen isotopic budget:
Implications for Holocene nitrogen cycling. Global Biogeochemical Cycles 16(4),
doi:10.1029/2001GB001856.
Bronk, D.A., 2002. Dynamics of DON. In: D.A. Hansell and C.A. Carlson (Editors),
Biogeochemistry of Marine Dissolved Organic Matter. Academic Press, New York, pp.
153-247.
95
References
Bronk, D.A., See, J.H., Bradley, P. and Killberg, L., 2007. DON as a source of bioavailable
nitrogen for phytoplankton. Biogeosciences 4, 283-296.
Calvert, S.E., Nielsen, B. and Fontugne, M.R., 1992. Evidence from nitrogen isotope ratios for
enhanced productivity during formation of Eastern Mediterranean sapropels. Nature 359,
223-225.
Canfield, D.E., Thamdrup, B. and Kristensen, E., 2005. The Nitrogen Cycle. In: A.J. Southward,
P.A. Tyler, C.M. Young and L.A. Fuiman (Editors), Aquatic Geomicrobiology. Advances
in Marine Biology. Elsevier Academic Press, Amsterdam, Boston, Heidelberg, London,
New York, Oxford, Paris, San Diego, San Francisco, Singapore, Sydney, Tokyo, pp. 205267.
Carpenter, E.J., Harvey, H.R., Fry, B. and Capone, D.G., 1997. Biogeochemical tracers of the
marine cyanobacterium Trichodesmium. Deep Sea Research I 44, 27-38.
Casciotti, K.L., Sigman, D.M., Hastings, M.G., Bohlke, J.K. and Hilkert, A., 2002. Measurement
of the oxygen isotopic composition of nitrate in seawater and freshwater using the
denitrifier method. Analytical Chemistry 74, 4905-4912.
Casciotti, K.L., Sigman, D.M. and Ward, B.B., 2003. Linking diversity and stable isotope
fractionation in ammonia-oxidizing bacteria. Geomicrobiology Journal 20, 335-353.
Casciotti, K.L. and Mc Ilvin, M.R., 2007. Isotopic analyses of nitrate and nitrite from reference
mixtures and application to Eastern Tropical North Pacific waters. Marine Chemistry 107,
184–201.
Casciotti, K.L., Trull, T.W., Glover, D.M. and Davies, D., 2008. Constraints on nitrogen cycling
at the subtropical North Pacific Station ALOHA from isotopic measurements of nitrate
and particulate nitrogen. Deep Sea Research II 55, 1661-1672.
Casciotti, K.L., 2009. Inverse kinetic isotope fractionation during bacterial nitrite oxidation.
Geochimica et Cosmochimica Acta 73(7), 2061-2076.
Cheddadi, R. and Rossignol-Strick, M., 1995. Eastern Mediterranean Quaternary Paleoclimates
from Pollen and Isotope Records of Marine Cores in the Nile Cone Area.
Paleoceanography 10(2), 291-300.
Cita, M.B., Vergnaud-Grazzini, C., Robert, C., Chamley, H., Ciaranfi, N. and d'Onofrio, S., 1977.
Paleoclimatic record of a long deep sea core from the eastern Mediterranean. Quaternary
Research 8(2), 205-235.
Clemens, S.C. and Prell, W.L., 1991. One million year record of summer monsoon winds and
continental aridity from the Owen Ridge (Site722), northwest Arabian Sea. In: W.D. Prell,
N. Niitsuma, K.C. Emeis and P.A. Meyers (Editors), Proceedings of the Ocean Drilling
Program, Scientific Results. Ocean Drilling Program, College Station, pp. 365-388.
Coban-Yildiz, Y., Altabet, M.A., Yilmaz, A. and Tugrul, S., 2006. Carbon and nitrogen isotopic
ratios of suspended particulate organic matter (SPOM) in the Black Sea water column.
Deep Sea Research II 53, 1875-1892.
96
References
Codispoti, L.A., Brandes, J.A., Christensen, J.P., Devol, A.H., Naqvi, S.W.A., Paerl, H.W. and
Yoshinari, T., 2001. The oceanic fixed nitrogen and nitrous oxide budgets: Moving targets
as we enter the anthropocene? Scientia Marina 65, 85-105.
Codispoti, L.A., 2007. An oceanic fixed nitrogen sink exceeding 400 Tg Na-1 vs the concept of
homeostasis in the fixed-nitrogen inventory. Biogeosciences 4(2), 233-253.
Cole, J.J., Honjo, S. and Erez, J., 1987. Benthic decomposition of organic-matter at a deep-water
site in the Panama Basin. Nature 327, 703-704.
Cowie, G.L. and Hedges, J.I., 1992. Sources and Reactivities of Amino-Acids in a Coastal
Marine-Environment. Limnology and Oceanography 37(4), 703-724.
Dähnke, K., Bahlmann, E. and Emeis, K.-C., 2008. Nitrogen turnover in the Elbe estuary –
assessment by means of stable nitrate isotopes. Limnology and Oceanography 53(4),
1504-1511.
Dähnke, K., Serna, A., Blanz, T. and Emeis, K.C., 2008. Sub-recent nitrogen-isotope trends in
sediments from Skagerrak (North Sea) and Kattegat: Changes in N-budgets and Nsources? Marine Geology 253, 92-98.
Dauwe, B. and Middelburg, J.J., 1998. Amino acids and hexosamines as indicators of organic
matter degradation state in North Sea sediments. Limnology and Oceanography 43(5),
782-798.
Dauwe, B., Middelburg, J.J., Herman, P.M.J. and Heip, C.H.R., 1999. Linking diagenetic
alteration of aminoacids and bulk organic matter reactivity. Limnology and Oceanography
44(7), 1809-1814.
De Corte, D., Yokokawa, T., Varela, M., Agogué, H. and Herndl, G.J., 2009. Spatial distribution
of Bacteria and Archaea and amoA gene copy numbers throughout the water column of
the Eastern Mediterranean Sea. The ISME Journal 3(2), 147-158.
De Lange, G.J., Thomson, J., Reitz, A., Slomp, C.P., Principato, M.S., Erba, E. and Corselli, C.,
2008. Synchronous basin-wide formation and redox-controlled preservation of a
Mediterranean sapropel. Nature Geoscience 1, doi:10.1038/ngeo283.
Dehairs, F., Chesselet, R. and Jedwab, J., 1980. Discrete suspended particles of barite and the
barium cycle in the open ocean. Earth and Planetary Science Letters 49(2), 528-550.
Dehairs, F., Goeyens, L., Stroobants, N., Bernard, P., Goyet, C., Poisson, A. and Chesselet, R.,
1990. On Suspended Barite and the Oxygen Minimum in the Southern Ocean. Global
Biogeochemical Cycles 40(1), 85-102.
Deutsch, C., Sigman, D.M., Thunell, R.C., Meckler, A.N. and Haug, G.H., 2004. Isotopic
constraints on glacial/interglacial changes in the oceanic nitrogen budget. Global
Biogeochemical Cycles 18(4), doi:10.1029/2003GB002189.
Diaz, F. and Raimbault, P., 2000. Nitrogen regeneration and dissolved organic nitrogen release
during spring in a NW Mediterranean coastal zone (Gulf of Lions): implications for the
estimation of new production. Marine Ecology Progress Series 197, 51-65.
Duce, R.A., LaRoche, J., Altieri, K., Arrigo, K.R., Baker, A.R., Capone, D.G., Cornell, S.,
Dentener, F., Galloway, J., Ganeshram, R.S., Geider, R.J., Jickells, T., Kuypers, M.M.,
97
References
Langlois, R., Liss, P.S., Liu, S.M., Middelburg, J.J., Moore, C.M., Nickovic, S., Oschlies,
A., Pedersen, T., Prospero, J., Schlitzer, R., Seitzinger, S., Sorensen, L.L., Uematsu, M.,
Ulloa, O., Voss, M., Ward, B. and Zamora, L., 2008. Impacts of atmospheric
anthropogenic nitrogen on the open ocean. Science 320, 893-897.
Dymond, J., Suess, E. and Lyle, M., 1992. Barium in deep-sea sediments: A geochemical proxy
for paleoproductivity. Paleoceanography 7, 163-181.
Emeis, K.C., Robertson, A.H.F., Richter, C. and others, 1996. Proceedings ODP, Initial Reports
160. Ocean Drilling Programm, College Station, TX.
Emeis, K.-C., Struck, U., Schulz, H.-M., Rosenberg, R., Bernasconi, S., Erlenkeuser, H.,
Sakamoto, T. and Martinez-Ruiz, F., 2000. Temperature and salinity variations of
Mediterranean Sea surface waters over the last 16,000 years from records of planktonic
stable oxygen isotopes and alkenone unsaturation ratios. Palaeogeography,
Palaeoclimatology, Palaeoecology 158, 259-280.
Emeis, K.C. and Weissert, H., 2009. Tethyan-Mediterranean organic carbon-rich sediments from
Mesozoic black shales to sapropels. Sedimentology 56, 247-266.
Eppley, R.W. and Peterson, B.J., 1979. Particulate organic-matter flux and planktonic new
production in the deep ocean. Nature 282, 677-680.
Fagerli, H., Legrand, M., Preunkert, S., Vestreng, V., Simpson, D. and Cerqueira, M., 2007.
Modeling historical long-term trends of sulfate, ammonium, and elemental carbon over
Europe: A comparison with ice core records in the Alps. Journal of Geophysical Research
- Atmospheres 112, doi:10.1029.2006JD008044
Farrell, G., Pedersen, T.F., Calvert, S.E. and Nielsen, B., 1995. Glacial-interglacial changes in
nutrient utilization in the equatorial Pacific Ocean. Nature 377, 514-517.
Francois, R., Altabet, M.A. and Burckle, L.H., 1992. Glacial to interglacial changes in surface
nitrate utilization in the Indian sector of the Southern Ocean as recorded by sediment
δ15N. Paleoceanography 7, 589-606.
Francois, R., Honjo, S., Manganini, S.J. and Ravizza, G.E., 1995. Biogenic Barium Fluxes to the
Deep-Sea - Implications for Paleoproductivity Reconstruction. Global Biogeochemical
Cycles 9(2), 289-303.
Francois, R., Altabet, M.A., Yu, E.F., Sigman, D.M., Bacon, M.P., Frank, M., Bohrmann, G.,
Bareille, G. and Labeyrie, L.D., 1997. Contribution of Southern Ocean surface-water
stratification to low atmospheric CO2 concentrations during the last glacial period. Nature
389, 929-935.
Freudenthal, T., Wagner, T., Wenzhöfer, F., Zabel, M. and Wefer, G., 2001. Early diagenesis of
organic matter from sediments of the eastern subtropical Atlantic: evidence from stable
nitrogen and carbon isotopes. Geochimica et Cosmochimica Acta 65, 1795-1808.
Fry, B., Jannasch, H.W., Molyneaux, S.J., Wirsen, C.O., Muramoto, J.A. and King, S., 1991.
Stable Isotope Studies of the Carbon, Nitrogen and Sulfur Cycles in the Black-Sea and the
Cariaco Trench. Deep Sea Research I 38, 1003-1019.
Fry, B., 2007. Stable Isotope Ecology. Springer-Verlag, Berlin, 308 pp.
98
References
Fuchsman, C.A., Murray, J.W. and Konovalov, S.K., 2008. Concentration and natural stable
isotope profiles of nitrogen species in the Black Sea. Marine Chemistry 111(1-2), 90-105.
Galloway, J.N., Aber, J.D., Erisman, J.W., Seitzinger, S.P., Howarth, R.W., Cowling, E.B. and
Cosby, B.J., 2003. The nitrogen cascade. Bioscience 53(4), 341-356.
Ganeshram, R.S., Pedersen, T.F., Calvert, S.E. and Francois, R., 2002. Reduced nitrogen fixation
in the glacial ocean inferred from changes in marine nitrogen and phosphorus inventories.
Nature 415, 156-159.
Garvin, J., Buick, R., Anbar, A.D., Arnold, G.L. and Kaufman, A.J., 2009. Isotopic Evidence for
an Aerobic Nitrogen Cycle in the Latest Archean. Science 323, 1045-1048.
Gat, J.R., Shemesh, A., Tziperman, E., Hecht, A., Georgopoulos, D. and Bastürk, O., 1996.
Journal of Geophysical Research 106(C3), 6441-6451.
Gaye-Haake, B., Lahajnar, N., Emeis, K.-C., Rixen, T., Unger, D., Schulz, H., Ramaswamy, V.,
Paropkari, A.L., Guptha, M.V.S. and Ittekkot, V., 2005. Stable nitrogen isotopic ratios of
sinking particles and sediments from the northern Indian Ocean. Marine Chemistry 96,
243-255.
Gaye, B., Wiesner, M.G. and Lahajnar, N., 2009. Nitrogen sources in the South China Sea as
discerned from stable nitrogen isotopic ratios in rivers, sinking particles, and sediments.
Marine Chemistry 114, 72-85.
Giunta, S., Negri, A., Morigi, C., Capotondi, L., Combourieu-Nebout, N., Emeis, K.C., Sangiorgi,
F. and Vigliotti, L., 2003. Coccolithophorid ecostratigraphy and multi-proxy
paleoceanographic reconstruction in the Southern Adriatic Sea during the last deglacial
time (Core AD91-17). Palaeogeography, Palaeoclimatology, Palaeoecology 190, 39-59.
Gouretski, V.V. and Koltermann, K.P., 2004. WOCE Global Hydrographic Climatology. Berichte
des Bundesamtes für Seeschifffahrt und Hydrographie 35, pp. 52.
Granger, J., Sigman, D.M., Needoba, J.A. and Harrison, P.J., 2004. Coupled nitrogen and oxygen
isotope fractionation of nitrate during assimilation by cultures of marine phytoplankton.
Limnology and Oceanography 49(5), 1763-1773.
Grasshoff, K., Ehrhardt, M., Kremling, K. and Anderson, L.G. (Editors), 1999. Methods of
Seawater Analysis. Verlag Chemie, Weinheim, 632 pp.
Gruber, N. and Sarmiento, J., 1997. Global patterns of nitrogen fixation and denitrification.
Global Biogeochemical Cycles 11, 235-266.
Gruber, N. and Galloway, J.N., 2008. An Earth-system perspective of the global nitrogen cycle.
Nature 451, 293-296.
Guptha, M.V.S., Naidu, P.D., Haake, B.G. and Schiebel, R., 2005. Carbonate and carbon
fluctuations in the Eastern Arabian Sea over 140 ka: Implications on productivity
changes? Deep Sea Research II 52, 1981-1993.
Hecky, R.E., Mopper, K., Kilham, P. and Degens, E.T., 1973. The Amino Acid and Sugar
Composition of Diatom Cell-Walls. Marine Biology 19, 323-331.
99
References
Higginson, M.J. and Altabet, M.A., 2004. Comment on "Sedimentary phosphorus record from the
Oman margin: New evidence of high productivity during glacial periods'' by F. Tamburini
et al.. Paleoceanography 19(2). doi:10.1029/2004PA001017.
Hoch, M.P., Fogel, M.L. and Kirchman, D.L., 1992. Isotope Fractionation Associated with
Ammonium Uptake by a Marine Bacterium. Limnology and Oceanography 37(7), 14471459.
Holmes, M.E., Muller, P.J., Schneider, R.R., Segl, M., Patzold, J. and Wefer, G., 1996. Stable
nitrogen isotopes in Angola Basin surface sediments. Marine Geology 134, 1-12.
Holmes, M.E., Müller, P.J., Segl, M. and Wefer, G., 1997. Reconstruction of past nutrient
utilization in the eastern Angola Basin based on sedimentary 15N/14N ratios.
Paleoceanography 12(4), 604-614.
Holmes, E., Lavik, G., Fischer, G., Segl, M., Ruhland, G. and Wefer, G., 2002. Seasonal
variability of δ15N in sinking particles in the Benguela upwelling region. Deep Sea
Research I 49, 377-394.
Honjo, S., Manganini, S.J. and Cole, J.J., 1982. Sedimentation of biogenic matter in the deep
ocean. Deep Sea Research I 29, 609-625.
Howell, M.W. and Thunell, R.C., 1992. Organic carbon accumulation in Bannock Basin:
Evaluating the role of productivity in the formation of eastern Mediterranean sapropels.
Marine Geology 103, 461-471.
Ibello, V., Cantoni, C., Cozzi, S. and Civitarese, G., in press. First basin-wide experimental results
on N2 fixation in the open Mediterranean Sea. Geophysical Research Letters.
Ingalls, A.E., Lee, C., Wakeham, S.G. and Hedges, J.I., 2003. The role of biominerals in the
sinking flux and preservation of amino acids in the Southern Ocean along 170 degrees W.
Deep Sea Research II 50, 713-738.
Ivanova, E., Schiebel, R., Singh, A.D., Schmiedl, G., Niebler, H.-S. and Hemleben, C., 2003.
Primary production in the Arabian Sea during the last 135.000 years. Palaeogeography,
Palaeoclimatology, Palaeoecology 197, 61-82.
Jaeschke, A., Hopmans, E.C., Wakeham, S.G., Schouten, S. and Sinninghe Damste, J.S., 2007.
The presence of ladderane lipids in the oxygen minimum zone of the Arabian Sea
indicates nitrogen loss through anammox. Limnology and Oceanography 52(2), 780-786.
Jenkyns, H.C., Grocke, D.R. and Hesselbo, S.P., 2001. Nitrogen isotope evidence for water mass
denitrification during the early Toarcian (Jurassic) oceanic anoxic event.
Paleoceanography 16(6), 593-603.
Jennerjahn, T.C. and Ittekkot, V., 1997. Organic Matter in sediments in the mangrove areas and
adjacent continental margins of Brazil: I. Amino acids and hexosamines. Oceanologica
Acta 20(2), 359-369.
Jennerjahn, T.C. and Ittekkot, V., 1999. Changes in organic matter from surface waters to
continental slope sediments off the Sao Francisco River, eastern Brazil. Marine Geology
161, 129-140.
100
References
Johannsen, A., Dähnke, K. and Emeis, K., 2008. Isotopic composition of nitrate in five German
rivers discharging into the North Sea. Organic Geochemistry 39, 1678-1689.
Jung, M., Ilmberger, J., Mangini, A. and Emeis, K.C., 1997. Why some Mediterranean sapropels
survived burn-down (and others did not). Marine Geology 141, 51-60.
Kaiser, K. and Benner, R., 2009. Biochemical composition and size distribution of organic matter
at the Pacific and Atlantic time-series stations. Marine Chemistry 113, 63-77.
Karl, D., Michaels, A., Bergman, B., Capone, D., Carpenter, E., Letelier, R., Lipschultz, F., Paerl,
H., Sigman, D. and Stal, L., 2002. Dinitrogen fixation in the world's oceans.
Biogeochemistry 57(1), 47-98.
Kemp, A.E.S., Pearce, R.B., Koizumi, I., Pike, J. and Rance, S.J., 1999. The role of mat-forming
diatoms in the formation of Mediterranean sapropels. Nature 398, 57-61.
Kendall, C., 1998. Tracing nitrogen sources and cycling in catchments. In: Kendall C &
McDonnell JJ (Editors) Isotope Tracers in Catchment Hydrology. Elsevier, Amsterdam,
pp. 521–576.
Kendall, C., Elliott, E.M. and Wankel, S.D., 2007. Tracing anthropogenic inputs of nitrogen to
ecosystems. In: K. Lajtha and R.H. Michener (Editors), Stable Isotopes in Ecology and
Environmental Science. Blackwell Scientific Publications, Oxford, pp. 375-449.
Knapp, A.N., Sigman, D. and Lipschultz, F., 2005. N-isotopic composition of dissolved organic
nitrogen and nitrate at the Bermuda Atlantic Time-Series study site. Global
Biogeochemical Cycles 19, doi:10.1029/2004GBC002320.
Knapp, A., DiFiore, P.J., Deutsch, C., Sigman, D.M. and Lipschultz, F., 2008. Nitrate isotopic
composition between Bermuda and Puerto Rico: Implications for N2 fixation in the
Atlantic Ocean Global Biogeochemical Cycles 22, doi: 10.1029/2007GB003107
Koppelmann, R., Böttger-Schnack, R., Möbius, J. and Weikert, H., 2009. Trophic relationships of
zooplankton in the eastern Mediterranean Sea based on stable isotope measurements.
Journal of Plankton Research 31, 669-686.
Kouvarakis, G., Mihalopoulos, N., Tselepides, A. and Stavrakakis, S., 2001. On the importance of
atmospheric inputs of inorganic nitrogen species on the propductivity of the eastern
Mediterranean Sea. Global Biogeochemical Cycles 15(4), 805-817.
Kress, N. and Herut, B., 2001. Spatial and seasonal evolution of dissolved oxygen and nutrients in
the Southern Levantine Basin (Eastern Mediterranean Sea): Chemical characterisation of
the water masses and inferences on the N: P ratios. Deep Sea Research I 48, 2347-2372.
Krom, M.D., Kress, N., Brenner, S. and Gordon, L.I., 1991. Phosphorus limitation of primary
production in the eastern Mediterranean Sea. Limnology and Oceanography 36, 424-432.
Krom, M.D., Groom, S. and Zohary, T., 2003. The Eastern Mediterranean. In: K.D. Black and
G.B. Shimmield (Editors), The Biogeochemistry of Marine Systems. Blackwell
Publishing, Oxford, pp. 91-122.
Krom, M.D., Herut, B. and Mantoura, R.F.C., 2004. Nutrient budget for the Eastern
Mediterranean: Implications for phosphorus limitation. Limnology and Oceanography
49(5), 1582-1592.
101
References
Krom, M.D., Woodward, E.M.S., Herut, B., Kress, N., Carbo, P., Mantoura, R.F.C., Spyres, G.,
Thingstad, T.F., Wassmann, P., Wexels-Riser, C., Kitidis, V., Law, C.S. and Zodiatis, G.,
2005a. Nutrient cycling in the south east Levantine basin of the eastern Mediterranean:
Results from a phosphorus starved system Deep Sea Research II 52, 2879–2896.
Krom, M.D., Thingstad, T.F., Brenner, S., Carbo, P., Drakopoulos, P., Fileman, T.W., Flaten,
G.A.F., Groom, S., Herut, B., Kitidis, V., Kress, N., Law, C.S., Liddicoat, M.I., Mantoura,
R.F.C., Pasternak, A., Pitta, P., Polychronaki, T., Psarra, S., Rassoulzadegan, F., Skjoldal,
E.F., Spyres, G., Tanaka, T., Tselepides, A., Wassmann, P., Wexels Riser, C., Woodward,
E.M.S., Zodiatis, G. and Zohary, T., 2005b. Summary and overview of the CYCLOPS P
addition Lagrangian experiment in the Eastern Mediterranean Sea. Deep Sea Research II
52, 3090-3108.
Kuhnt, T., Schmiedl, G., Ehrmann, W., Hamann, Y. and Hemleben, C., 2007. Deep-sea
ecosystem variability of the Aegean Sea during the past 22 kyr as revealed by Benthic
Foraminifera. Marine Micropaleontology 64, 141-162.
Kuhnt, T., Schmiedl, G., Ehrmann, W., Hamann, Y. and Andersen, N., 2008. Stable isotopic
composition of Holocene benthic foraminifers from the Eastern Mediterranean Sea: Past
changes in productivity and deep water oxygenation. Palaeogeography,
Palaeoclimatology, Palaeoecology 268,106-115.
Kullenberg, B., 1952. On the salinity of water contained in marine sediments. Meddelanden fran
Oceanografiska Institutet i Göteborg 21, 1-38.
Kuypers, M.M.M., van Breugel, Y., Schouten, S., Erba, E. and Damste, J.S.S., 2004. N2-fixing
cyanobacteria supplied nutrient N for Cretaceous oceanic anoxic events. Geology 32, 853856.
Lahajnar, N., Wiesner, M.G. and Gaye, B., 2007. Fluxes of amino acids and hexosamines to the
deep South China Sea. Deep Sea Research I 54, 2120-2144.
Lascaratos, A., Roether, W., Nittis, K. and Klein, B., 1999. Recent changes in deep water
formation and spreading in the eastern Mediterranean Sea: a review. Progress in
Oceanography 44, 5–36.
Lee, C. and Cronin, C., 1982. The vertical flux of particulate organic nitrogen in the sea:
decomposition of amino acids in the Peru upwelling area and the equatorial Atlantic.
Journal of Marine Research 40, 227-251.
Lee, C. and Cronin, C., 1984. Particulate amino acids in the sea: Effects of primary productivity
and biological decomposition. Journal of Marine Research 42, 1075-1097.
LeGrande, N.A. and Schmidt, G.A., 2006. Global gridded data set of the oxygen isotopic
composition in seawater. Geophysical Research Letters 33, doi:10.1029/2006GL026011.
Lehmann, M.F., Bernasconi, S.M., Barbieri, A. and McKenzie, J.A., 2002. Preservation of
organic matter and alteration of its carbon and nitrogen isotope composition during
simulated and in situ early sedimentary diagenesis. Geochimica et Cosmochimica Acta 66,
3573-3584.
102
References
Lehmann, M.F., Reichert, P., Bernasconi, S.M., Barbieri, A. and McKenzie, J.A., 2003.
Modelling nitrogen and oxygen isotope fractionation during denitrification in a lacustrine
redox-transition zone. Geochimica et Cosmochimica Acta 67, 2529-2542.
Lehmann, M.F., Sigman, D.M., McCorkle, D.C., Brunelle, B.G., Hoffmann, S., Kienast, M.,
Cane, G. and Clement, J., 2005. The origin of the deep Bering Sea nitrate deficit Constraints from the nitrogen and oxygen isotopic composition of water-column nitrate
and
benthic
nitrate
fluxes.
Global
Biogeochemical
Cycles
19,
doi:10.1029/2005GB002508.
Leuschner, D.C. and Sirocko, F., 2003. Orbital insolation forcing of the Indian Monsoon - a motor
for global climate changes? Palaeogeography, Palaeoclimatology, Palaeoecology 197, 8395.
Li, Q.P., Hansell, D.A. and Zhang, J.Z., 2008. Underway monitoring of nanomolar nitrate plus
nitrite and phosphate in oligotrophic seawater. Limnology and Oceanography - Methods,
6: 319-326.
Libes, S.M., 1992. An Introduction to Marine Biogeochemistry. John Wiley & Sons, Inc., New
York, Chichester, Brisbane, Toronto, Singapore, 734 pp.
Macko, S.A. and Estep, M.L.F., 1984. Microbial alteration of stable nitrogen and carbon isotopic
compositions of organic matter. Organic Geochemistry 6, 787-790.
Mahaffey, C., Michaels, A. and Capone, D.G., 2005. The conundrum of marine nitrogen fixation.
American Journal of Science 305, 546–595.
Malanotte-Rizzolli, P. and Bergamasco, A., 1989. The circulation of the eastern Mediterrean, Part
1. Oceanologica Acta 12, 335-351.
Mara, P., Mihalopoulos, N., Gogou, A., Dähnke, K., Schlarbaum, T., Emeis, K.-C. and Krom, M.,
2009. Isotopic composition of nitrate in wet and dry atmospheric deposition on Crete in
the
eastern
Mediterranean
Sea.
Global
Biogeochemical
Cycles
23,
doi:10.1029/2008gb003395
Mariotti, A., Germon, J.C., Hubert, P., Kaiser, P., Letolle, R., Tardieux, A. and Tardieux, P.,
1981. Experimental determination of nitrogen kinetic isotope fractionation: some
principles; illustration for the denitrification and nitrification processes. Plant and Soil 62,
413-430.
Mariotti, A., 1984. Natural N-15 abundance measurements and atmospheric nitrogen standard
calibration. Nature 311, 251-252.
Martin, J.H., Gordon, R.M. and Fitzwater, S.E., 1990. Iron in Antarctic Waters. Nature 345, 156158.
Mayer, B., Boyer, E.W., Goodale, C., Jaworski, N.A., Van Breemen, N., Howarth, R.W.,
Seitzinger, S., Billen, G., Lajtha, L.J., Nosal, M. and Paustian, K., 2002. Sources of nitrate
in rivers draining sixteen watersheds in the northeastern US: Isotopic constraints.
Biogeochemistry 57(1), 171-197.
McCarthy, J.J., Yilmaz, A., Coban-Yildiz, Y. and Nevins, J.L., 2007. Nitrogen cycling in the
offshore waters of the Black Sea. Estuarine, Coastal and Shelf Science 74, 493-514.
103
References
McClelland, J.W. and Valiela, I., 1998. Linking nitrogen in estuarine producers to land-derived
sources. Limnology and Oceanography 43(4), 577-585.
McCoy, F.W., 1974. Late Quaternary sediments in the Eastern Mediterranean Sea. Ph.D. thesis,
Harvard University, Cambridge, MA.
Meador, T.B., Aluwihare, L.I. and Mahaffey, C., 2007. Isotopic heterogeneity and cycling of
organic nitrogen in the oligotrophic ocean. Limnology and Oceanography 52 (3), 934-947.
Mercone, D., Thomson, J., Abu-Zied, R.H., Croudace, I.W. and Rohling, E.J., 2001. Highresolution geochemical and micropalaeontological profiling of the most recent eastern
Mediterranean sapropel. Marine Geology 177, 25-44.
Meyers, P.A. and Bernasconi, S.M., 2005. Carbon and nitrogen isotope excursions in midPleistocene sapropels from the Tyrrhenian Basin: Evidence for climate-induced increases
in microbial primary production. Marine Geology 220, 41-58.
Milder, J.C. and Montoya, J.P., 1999. Carbon and nitrogen stable isotope ratios at Sites 969 and
974: interpreting spatial gradients in sapropel properties. In: R. Zahn, M. Comas and A.
Klaus (Editors), Proceedings ODP, Scientific Results, 161. Ocean Drilling Programm,
College Station, pp. 401-411.
Minagawa, M. and Wada, E., 1986. Nitrogen isotope ratios of red tide organisms in the East
China Sea: A characterization of biological nitrogen fixation. Marine Chemistry 19, 245249.
Montoya, J., Carpenter, E. and Capone, D., 2002. Nitrogen fixation and nitrogen isotope
abundance in zooplankton of the oligotrophic North Atlantic. Limnology and
Oceanography 47, 1617– 1628.
Moodley, L., Middleburg, J.J., Herman, P.M.J., Soetaert, K. and de Lange, G.J., 2005.
Oxygenation and organic-matter preservation in marine sediments: Direct experimental
evidence from ancient organic carbon–rich deposits. Geology 33, 889-892.
Müller, P.J. and Suess, E., 1979. Productivity, sedimentation rate, and sedimentary organic matter
in the oceans-I. Organic carbon preservation. Deep Sea Research I 26, 1347-1362.
Murat, A. and Got, H., 2000. Organic carbon variations of the eastern Mediterranean Holocene
sapropel: a key for understanding formation processes. Palaeogeography,
Palaeoclimatology, Palaeoecology 158, 241- 257.
Naqvi, S.W.A., 1994. Denitrification Processes in the Arabian Sea. Proceedings of the Indian
Academy of Sciences-Earth and Planetary Sciences 103(2), 279-300.
Naqvi, S.W.A., Yoshinari, T., Jayakumar, D.A., Altabet, M.A., Narvekar, P.V., Devol, A.H.,
Brandes, J.A. and Codispoti, L.A., 1998. Budgetary and biogeochemical implications of
N2O isotope signatures in the Arabian Sea. Nature 394, 462-464.
Pantoja, S., Repeta, D.J., Sachs, J.P. and Sigman, D.M., 2002. Stable isotope constraints on the
nitrogen cycle of the Mediterranean Sea water column. Deep Sea Research I 49, 16091621.
104
References
Pichevin, L., Martinez, P., Bertrand, P., Schneider, R. and Emeis, K.-C., 2005. Nitrogen cycling
on the Namibian shelf and slope over the last two climatic cycles: Local and global
forcings. Paleoceanography 20, doi:10.1029/2006GB002852.
Pichevin, L., Bard, E., Martinez, P. and Billy, I., 2007. Evidence of ventilation changes in the
Arabian Sea during the late Quarternary: Implication for denitrification and nitrous oxide
emission. Global Biogeochemical Cycles 21, doi:10.1029/2006GB002852.
Pierre, C., 1999. The oxygen and carbon isotope distribution in the Mediterranean water masses.
Marine Geology 153, 41-55.
Prell, W.D., Niitsuma, N. and others, 1989. Proceedings ODP, Initial Reports 117. Ocean Drilling
Program, College Station, TX.
Preunkert S., Wagenbach D. and Legrand M., 2003. A seasonally resolved alpine ice core record
of nitrate: Comparison with anthropogenic inventories and estimation of preindustrial
emissions of NO in Europe. Journal of Geophysical Research 108,
doi:10.1029/2003JD003475.
Redfield, A.C., 1934. On the proportions of organic derivatives in sea-water and their relation to
the composition of plankton. In: R. J. Daniel (Editor), James Johnstone Memorial
Volume, University Press of Liverpool, pp. 176-192.
Rees, A.P., Law, C.S. and Woodward, E.M.S., 2006. High rates of nitrogen fixation during an insitu phosphate release experiment in the Eastern Mediterranean Sea. Geophysical
Research Letters 33, doi:10.1029/2006GL025791.
Ren, H., Sigman, D., Meckler, A.N., Plessen, B., Robinson, R.S., Rosenthal, Y. and Haug, G.H.,
2009. Foraminiferal Isotope Evidence of Reduced Nitrogen Fixation in the Ice Age
Atlantic Ocean. Science 323, 244-248.
Reschke, S., 1999. Biogeochemie der Schwebstofffracht der Donau und deren Einfluß auf das
Sedimentationsgeschehen im nordwestlichen Schwarzen Meer. Ph.D. thesis, Universität
Hamburg, Hamburg, 109 pp.
Ribera d´Alcala, M., Civitarese, G., Conversano, F. and Lavezza, R., 2003. Nutrient ratios and
fluxes hint at overlooked processes in the Mediterranean Sea. Journal of Geophysical
Research 108(C9), doi:10.1029/2002JC001650.
Rixen, T., Haake, B. and Ittekkot, V., 2000. Sedimentation in the western Arabian Sea: The role
of water mass advection, coastal and open ocean upwelling. Deep Sea Research II 47,
2155-2178.
Roether, W., Manca, B.B., Klein, B., Bregant, D., Georgopoulos, D., Beitzel, V., Kovacevic, V.
and Luchetta, A., 1996. Recent changes in Eastern Mediterranean deep waters. Science
271, 333-335.
Rohling, E.J., 1994. Review and new aspects concerning the formation of eastern Mediterranean
sapropels. Marine Geology 122, 1-28.
Rohling, E.J., Hopmans, E.C. and Damste, J.S.S., 2006. Water column dynamics during the last
interglacial anoxic event in the Mediterranean (sapropel S5). Paleoceanography 21(2),
doi:10.1029/2005PA001237.
105
References
Rossignol-Strick, M., Nesteroff, W., Olive, P. and Vergnaudgrazzini, C., 1982. After the Deluge Mediterranean Stagnation and Sapropel Formation. Nature 295, 105-110.
Rossignol-Strick, M., 1983. African Monsoons, an Immediate Climate Response to Orbital
Insolation. Nature 304, 46-49.
Rossignol-Strick, M., 1985. Mediterranean Quaternary sapropels, an immediate response of the
African monsoon to variation of insolation. Palaeogeography, Palaeoclimatology,
Palaeoecology 49, 237-263.
Ryan, W.B.F., 1972. Stratigraphy of Late Quarternary sediments in the eastern Mediterranean. In:
D.J. Stanley (Editor), The Mediterranean Sea: a Natural Sedimentation Laboratory.
Dowden, Hutchinson and Ross, Stroudsberg, Pa, pp. 149-169.
Sachs, J.P. and Repeta, D.J., 1999. Oligotrophy and nitrogen fixation during Eastern
Mediterranean sapropel events. Science 286, 2485-2488.
Saino, T. and Hattori, A., 1980. N-15 Natural Abundance in Oceanic Suspended Particulate
Matter. Nature 283, 752-754.
Schlarbaum, T., Dähnke, K., Bahlmann, E. and Emeis, K.-C., in revision. Dissolved organic
nitrogen in the Elbe River and estuary: Results of nitrogen isotope investigations. Marine
Chemistry.
Schulz, H., von Rad, U. and Erlenkeuser, H., 1998. Correlation between Arabian Sea and
Greenland climate oscillations of the past 110000 years. Nature 393, 54-57.
Sigman, D.M., Altabet, M.A., Michener, R., McCorkle, D.C., Fry, B. and Holmes, R.M., 1997.
Natural abundance-level measurement of the nitrogen isotopic composition of oceanic
nitrate: an adaptation of the ammonia diffusion method. Marine Chemistry 57, 227-242.
Sigman, D.M., Altabet, M.A., Francois, R., McCorkle, D.C. and Gaillard, J.F., 1999. The isotopic
composition of diatom-bound nitrogen in Southern Ocean sediments. Paleoceanography
14(2), 118-134.
Sigman, D.M., Altabet, M.A., McCorkle, D.C., Francois, R. and Fischer, G., 2000. The delta N15 of nitrate in the Southern Ocean: Nitrogen cycling and circulation in the ocean interior.
Journal of Geophysical Research-Oceans 105(C8), 19599-19614.
Sigman, D.M., Casciotti, K.L., Andreani, M., Barford, C., Galanter, M. and Bohlke, J.K., 2001. A
bacterial method for the nitrogen isotopic analysis of nitrate in seawater and freshwater.
Analytical Chemistry 73, 4145-4153.
Sigman, D.M., Granger, J., DiFiore, P.J., Lehmann, M.M., Ho, R., Cane, G. and van Geen, A.,
2005. Coupled nitrogen and oxygen isotope measurements of nitrate along the eastern
North Pacific margin. Global Biogeochemical Cycles 19(4), doi:10.1029/2005GB002458.
Sigman, D.M., DiFiore, P.J., Hain, M.P., Deutsch, C., Wang, Y., Karl, D.M., Knapp, A.N.,
Lehmann, M.F. and Pantoja, S., 2009. The dual isotopes of deep nitrate as a constraint on
the cycle and budget of oceanic fixed nitrogen Deep Sea Research I 56, 1419-1439.
Silfer, J.A., Engel, M.H. and Macko, S.A., 1992. Kinetic Fractionation of Stable Carbon and
Nitrogen Isotopes During Peptide-Bond Hydrolysis - Experimental-Evidence and
Geochemical Implications. Chemical Geology 101, 211-221.
106
References
Sirocko, F. and Ittekkot, V., 1992. Organic-Carbon Accumulation Rates in the Holocene and
Glacial Arabian Sea - Implications for O2-Consumption in the Deep-Sea and Atmospheric
CO2 Variations. Climate Dynamics 7(4), 167-172.
Slomp, C.P., Thomson, J. and de Lange, G.J., 2002. Enhanced regeneration of phosphorus during
formation of the most recent eastern Mediterranean sapropel (S1). Geochimica et
Cosmochimica Acta 66, 1171-1184.
Sprent, J.I. and Sprent, P., 1990. Nitrogen Fixing Organisms: Pure and Applied Aspects.
Chapman and Hall, London, 256 pp.
Struck, U., Emeis, K.-C., Rau, G.H., Voss, M., Krom, M.D. and Rau, G.H., 2001. Biological
productivity during sapropel S5 formation in the eastern Mediterranean Sea - Evidence
from stable isotopes of nitrogen and carbon. Geochimica et Cosmochimica Acta 65, 32413258.
Suthhof, A., Jennerjahn, T.C., Schafer, P. and Ittekkot, V., 2000. Nature of organic matter in
surface sediments from the Pakistan continental margin and the deep Arabian Sea: amino
acids. Deep Sea Research II 47, 329-351.
Suthhof, A., Ittekkot, V. and Gaye-Haake, B., 2001. Millennial-scale oscillation of denitrification
intensity in the Arabian Sea during the late Quaternary and its potential influence on
atmospheric N2O and global climate. Global Biogeochemical Cycles 15(3), 637-649.
Sweeney, R.E. and Kaplan, I.R., 1980. Natural abundances of 15 N as a source indicator for nearshore marine sedimentary and dissolved nitrogen. Marine Chemistry 9, 81-90.
Tamburini, F., Föllmi, K.B., Adatte, T., Bernasconi, S.M. and Steinmann, P., 2003. Sedimentary
phosphorus record from the Oman margin: New evidence of high productivity during
glacial periods. Paleoceanography 18(1), doi:10.1029/2000PA000616.
Thingstad, T.F. and Rassoulzadegan, F., 1999. Conceptual models for the biogeochemical role of
the photic zone microbial food web, with particular reference to the Mediterranean Sea.
Progress in Oceanography 44, 271–286.
Thingstad, T.F., Krom, M.D., Mantoura, R.F.C., Flaten, G.A.F., Groom, S., Herut, B., Kress, N.,
Law, C.S., Pasternak, A., Pitta, P., Psarra, S., Rassoulzadegan, F., Tanaka, T., Tselepides,
A., Wassmann, P., Woodward, E.M.S., Wexels Riser, C., Zodiatis, G. and Zohary, T.,
2005. Nature of phosphorus limitation in the ultraoligotrophic Eastern Mediterranean.
Science 309, 1068-1071.
Thomson, J., Higgs, N.C., Wilson, T.R.S., Croudace, I.W., De Lange, G.J. and Van Santvoort,
P.J.M., 1995. Redistribution and geochemical behaviour of redox-sensitive elements
around S1, the most recent eastern Mediterranean sapropel. Geochimica et Cosmochimica
Acta 59, 3487-3501.
Thomson, J., Mercone, D., De Lange, G.J. and Van Santvoort, P.J.M., 1999. Review of recent
advances in the interpretation of eastern Mediterranean sapropel S1 from geochemical
evidence. Marine Geology 153, 77-89.
Thunell, R.C., Sigman, D.M., Muller-Karger, F., Astor, Y. and Varela, R., 2004. Nitrogen isotope
dynamics of the Cariaco Basin, Venezuela. Global Biogeochemical Cycles 18(3),
doi:10.1029/2003GB002185.
107
References
Tissot, B.P. and Welte, D.H., 1978. Petroleum Formation and Occurrence. Springer-Verlag,
Berlin Heidelberg New York, 538 pp.
Triantaphyllou, M.V., Antonarakou, A., Kouli, K., Dimiza, M., Kontakiotis, G., Papanikolaou,
M.D., Ziveri, P., Mortyn, P.G., Lianou, V., Lykousis, V. and Dermitzakis, M.D., 2009.
Late Glacial-Holocene ecostratigraphy of the south-eastern Aegean Sea, based on
plankton and pollen assemblages. Geo-Marine Letters, doi:10.1007/s00367-009-0139-5.
Tyrrell, T., 1999. The relative influences of nitrogen and phosphorus on oceanic primary
production. Nature 400, 525-531.
Unger, D., Gaye-Haake, B., Neumann, K., Gebhardt, A.C. and Ittekkot, V., 2005.
Biogeochemistry of suspended and sedimentary material in the Ob and Yenisei rivers and
Kara Sea: amino acids and amino sugars. Continental Shelf Research 25, 437-460.
Voss, M., Nausch, G. and Montoya, J.P., 1997. Nitrogen stable isotope dynamics in the central
Baltic Sea: influence of deep-water renewal on the N-cycle changes. Marine Ecology
Progress Series 158, 11-21.
Voss, M., Dippner, J.W. and Montoya, J.P., 2001. Nitrogen isotope patterns in the oxygendeficient waters of the Eastern Tropical North Pacific Ocean. Deep Sea Research I 48,
1905-1921.
Voss, M., Emeis, K.C., Hille, S., Neumann, T. and Dippner, J.W., 2005. Nitrogen cycle of the
Baltic Sea from an isotopic perspective. Global Biogeochemical Cycles 19(3),
doi:10.1029/2004GB002338.
Voss, M., Deutsch, B., Elmgren, R., Humborg, C., Kuuppo, P., Pastuszak, M., Rolff, C. and
Schulte, U., 2006. Source identification of nitrate by means of isotopic tracers in the
Baltic Sea catchments. Biogeosciences 3, 663-676.
Wallmann, K., 2003. Feedbacks between oceanic redox states and marine productivity: A model
perspective focused on benthic phosphorus cycling. Global Biogeochemical Cycles 17(3),
doi:10.1029/2002GB001968.
Wankel, S.D., Kendall, C., Francis, C.A. and Paytan, A., 2006. Nitrogen sources and cycling in
the San Francisco Bay Estuary: A nitrate dual isotopic composition approach. Limnology
and Oceanography 51(4), 1654-1664.
Wankel, S.D., Chen, Y., Kendall, C., Post, A.F. and Paytan, A., 2009. Sources of aerosol nitrate to
the Gulf of Aqaba: Evidence from δ15N and δ18O of nitrate and trace metal chemistry.
Marine Chemistry, in press, corrected proof.
Ward, B.B., Devol, A.H., Rich, J.J., Chang, B.X., Bulow, S.E., Naik, H., Pratihary, A. and
Jayakumar, A., 2009. Denitrification as the dominant nitrogen loss process in the
Arabian Sea. Nature 461, 78-81.
Warnken, C., 2003. Biogeochemie von Schwebstoffen, Sinkstoffen und Sedimenten im IerapetraTief (östliches Mittelmeer). M.Sc. thesis, Universität Hamburg, Hamburg, 80 pp.
108
References
Westley, M.B., Yamagishi, H., Popp, B.N. and Yoshida, N., 2006. Nitrous oxide cycling in the
Black Sea inferred from stable isotope and isotopomer distributions. Deep Sea Research II
53, 1802-1816.
Wüst, G., 1961. On the vertical circulation of the Mediterranean Sea. Journal of Geophysical
Research 66, 3261-3271.
Wyrtki, K., 1971. Oceanographic Atlas of the International Indian Ocean Expedition. National
Science Foundation Publication, Washington DC, 531 pp.
Yool, A., Martin, A.P., Fernandez, C. and Clark, D.R., 2007. The significance of nitrification for
oceanic new production. Nature 447, 999-1002.
York, J.K., Tomasky, G., Valiela, I. and Repeta, D.J., 2007. Stable isotopic detection of
ammonium and nitrate assimilation by phytoplankton in the Waquoit Bay estuarine
system. Limnology and Oceanography 52, 144–155.
Zahn, R. and Pedersen, T.F., 1991. Late Pleistocene evolution of surface and mid-depth
hydrography at the Oman margin: Planktonic and benthic isotope records at Site 724. In:
W.D. Prell, N. Niitsuma, K.C. Emeis and P.A. Meyers (Editors), Proceedings ODP,
Scientific Results. College Station, TX, pp. 291-308.
Zohary, T. and Robarts, R.D., 1998. Experimental study of microbial P limitation in the eastern
Mediterranean. Limnology and Oceanography 43(3), 387-395.
Zohary, T., Herut, B., Krom, M., Mantoura, R.F.C., Pitta, P., Psarra, S., Rassoulzadegan, F.,
Stambler, N., Tanaka, T., Thingstad, T.F. and Woodward, E.M.S., 2005. P-limited
bacteria but N and P co-limited phytoplankton in the Eastern Mediterranean - a
microcosm experiment. Deep Sea Research II 52, 3011-3023.
109
Curriculum Vitae
Jürgen Henning Möbius
Born:
14.04. 1978 in Hamburg
Nationality:
German
Home address:
Lange Reihe 76, 20099 Hamburg
Education
Since
April 2006
Ph.D. student, Institut für Biogeochemie und Meereschemie,
Universität Hamburg
Thesis subject: Nitrogen cycling in the past and in the present
Mediterranean Sea and Arabian Sea – Implications from stable isotope
studies
March 2006
Graduation (Diploma)
Diploma-Thesis: Particle sedimentation during the 1998/1999 El
Niño event in the South China Sea: amino acids and amino
sugars
2002
Exchange student (spring semester) in geology at
“Universidad de Granada”, Granada, Spain
1999 - 2006
Studies of Geology and Paleontology (Diploma),
Universität Hamburg.
Emphasis on Biogeochemistry and Soil Science
1997
110
Abitur at Gymnasium Winsen
List of Publications:
Wieland, A., Pape, T., Möbius, J., Klock, J.H. and Michaelis, W., 2008: Carbon pools and
isotopic trends in a hypersaline cyanobacterial mat. Geobiology 6(2), 171-186.
Koppelmann, R., Böttger-Schnack, R., Möbius, J. and Weikert, H., 2009: Trophic relationships of
zooplankton in the eastern Mediterranean based on stable isotope measurements. Journal of
Plankton Research 31(6), 669-686.
Zonneveld, K.A.F., Chen, L., Möbius, J., Mahmoud, M.S., 2009: Environmental significance of
dinoflagellate cysts from the proximal part of the Po-river discharge plume (off southern Italy,
Eastern Mediterranean). Journal of Sea Research 62(4), 189-213.
Möbius, J., Lahajnar, N. and Emeis, K.-C., 2010: Diagenetic control of nitrogen isotope ratios in
Holocene sapropels and recent sediments from the Eastern Mediterranean Sea. Biogeosciences
Discussions 7, 1131-1165.
Emeis, K.-C., Mara, P., Schlarbaum, T., Möbius, J., Dähnke, K., Struck, U., Mihalopoulos, N. and
Krom, M.,: N-isotope ratios of nitrate, dissolved organic nitrogen and particulate nitrogen in the
Eastern Mediterranean Sea. Journal of Geophysical Research – Biogeochemistry (under review)
Möbius, J., Gaye, B., Lahajnar, N., Bahlmann, E. and Emeis, K.-C.,: Influence of diagenesis on
sedimentary δ15N in the Arabian Sea over the last 130 kyr. Submitted to Marine Geology
111